Application of a predictive Bayesian model to environmental accounting.

PubMed

Anex, R P; Englehardt, J D

2001-03-30

Environmental accounting techniques are intended to capture important environmental costs and benefits that are often overlooked in standard accounting practices. Environmental accounting methods themselves often ignore or inadequately represent large but highly uncertain environmental costs and costs conditioned by specific prior events. Use of a predictive Bayesian model is demonstrated for the assessment of such highly uncertain environmental and contingent costs. The predictive Bayesian approach presented generates probability distributions for the quantity of interest (rather than parameters thereof). A spreadsheet implementation of a previously proposed predictive Bayesian model, extended to represent contingent costs, is described and used to evaluate whether a firm should undertake an accelerated phase-out of its PCB containing transformers. Variability and uncertainty (due to lack of information) in transformer accident frequency and severity are assessed simultaneously using a combination of historical accident data, engineering model-based cost estimates, and subjective judgement. Model results are compared using several different risk measures. Use of the model for incorporation of environmental risk management into a company's overall risk management strategy is discussed.

Final Report, DOE Early Career Award: Predictive modeling of complex physical systems: new tools for statistical inference, uncertainty quantification, and experimental design

DOE Office of Scientific and Technical Information (OSTI.GOV)

Marzouk, Youssef

Predictive simulation of complex physical systems increasingly rests on the interplay of experimental observations with computational models. Key inputs, parameters, or structural aspects of models may be incomplete or unknown, and must be developed from indirect and limited observations. At the same time, quantified uncertainties are needed to qualify computational predictions in the support of design and decision-making. In this context, Bayesian statistics provides a foundation for inference from noisy and limited data, but at prohibitive computional expense. This project intends to make rigorous predictive modeling *feasible* in complex physical systems, via accelerated and scalable tools for uncertainty quantification, Bayesianmore » inference, and experimental design. Specific objectives are as follows: 1. Develop adaptive posterior approximations and dimensionality reduction approaches for Bayesian inference in high-dimensional nonlinear systems. 2. Extend accelerated Bayesian methodologies to large-scale {\\em sequential} data assimilation, fully treating nonlinear models and non-Gaussian state and parameter distributions. 3. Devise efficient surrogate-based methods for Bayesian model selection and the learning of model structure. 4. Develop scalable simulation/optimization approaches to nonlinear Bayesian experimental design, for both parameter inference and model selection. 5. Demonstrate these inferential tools on chemical kinetic models in reacting flow, constructing and refining thermochemical and electrochemical models from limited data. Demonstrate Bayesian filtering on canonical stochastic PDEs and in the dynamic estimation of inhomogeneous subsurface properties and flow fields.« less

[Bayesian geostatistical prediction of soil organic carbon contents of solonchak soils in nor-thern Tarim Basin, Xinjiang, China.

PubMed

Wu, Wei Mo; Wang, Jia Qiang; Cao, Qi; Wu, Jia Ping

2017-02-01

Accurate prediction of soil organic carbon (SOC) distribution is crucial for soil resources utilization and conservation, climate change adaptation, and ecosystem health. In this study, we selected a 1300 m×1700 m solonchak sampling area in northern Tarim Basin, Xinjiang, China, and collected a total of 144 soil samples (5-10 cm). The objectives of this study were to build a Baye-sian geostatistical model to predict SOC content, and to assess the performance of the Bayesian model for the prediction of SOC content by comparing with other three geostatistical approaches [ordinary kriging (OK), sequential Gaussian simulation (SGS), and inverse distance weighting (IDW)]. In the study area, soil organic carbon contents ranged from 1.59 to 9.30 g·kg -1 with a mean of 4.36 g·kg -1 and a standard deviation of 1.62 g·kg -1 . Sample semivariogram was best fitted by an exponential model with the ratio of nugget to sill being 0.57. By using the Bayesian geostatistical approach, we generated the SOC content map, and obtained the prediction variance, upper 95% and lower 95% of SOC contents, which were then used to evaluate the prediction uncertainty. Bayesian geostatistical approach performed better than that of the OK, SGS and IDW, demonstrating the advantages of Bayesian approach in SOC prediction.

Impact assessment of extreme storm events using a Bayesian network

USGS Publications Warehouse

den Heijer, C.(Kees); Knipping, Dirk T.J.A.; Plant, Nathaniel G.; van Thiel de Vries, Jaap S. M.; Baart, Fedor; van Gelder, Pieter H. A. J. M.

2012-01-01

This paper describes an investigation on the usefulness of Bayesian Networks in the safety assessment of dune coasts. A network has been created that predicts the erosion volume based on hydraulic boundary conditions and a number of cross-shore profile indicators. Field measurement data along a large part of the Dutch coast has been used to train the network. Corresponding storm impact on the dunes was calculated with an empirical dune erosion model named duros+. Comparison between the Bayesian Network predictions and the original duros+ results, here considered as observations, results in a skill up to 0.88, provided that the training data covers the range of predictions. Hence, the predictions from a deterministic model (duros+) can be captured in a probabilistic model (Bayesian Network) such that both the process knowledge and uncertainties can be included in impact and vulnerability assessments.

Quantifying model-structure- and parameter-driven uncertainties in spring wheat phenology prediction with Bayesian analysis

DOE PAGES

Alderman, Phillip D.; Stanfill, Bryan

2016-10-06

Recent international efforts have brought renewed emphasis on the comparison of different agricultural systems models. Thus far, analysis of model-ensemble simulated results has not clearly differentiated between ensemble prediction uncertainties due to model structural differences per se and those due to parameter value uncertainties. Additionally, despite increasing use of Bayesian parameter estimation approaches with field-scale crop models, inadequate attention has been given to the full posterior distributions for estimated parameters. The objectives of this study were to quantify the impact of parameter value uncertainty on prediction uncertainty for modeling spring wheat phenology using Bayesian analysis and to assess the relativemore » contributions of model-structure-driven and parameter-value-driven uncertainty to overall prediction uncertainty. This study used a random walk Metropolis algorithm to estimate parameters for 30 spring wheat genotypes using nine phenology models based on multi-location trial data for days to heading and days to maturity. Across all cases, parameter-driven uncertainty accounted for between 19 and 52% of predictive uncertainty, while model-structure-driven uncertainty accounted for between 12 and 64%. Here, this study demonstrated the importance of quantifying both model-structure- and parameter-value-driven uncertainty when assessing overall prediction uncertainty in modeling spring wheat phenology. More generally, Bayesian parameter estimation provided a useful framework for quantifying and analyzing sources of prediction uncertainty.« less

BAYESIAN METHODS FOR REGIONAL-SCALE EUTROPHICATION MODELS. (R830887)

EPA Science Inventory

We demonstrate a Bayesian classification and regression tree (CART) approach to link multiple environmental stressors to biological responses and quantify uncertainty in model predictions. Such an approach can: (1) report prediction uncertainty, (2) be consistent with the amou...

A Bayesian prediction model between a biomarker and the clinical endpoint for dichotomous variables.

PubMed

Jiang, Zhiwei; Song, Yang; Shou, Qiong; Xia, Jielai; Wang, William

2014-12-20

Early biomarkers are helpful for predicting clinical endpoints and for evaluating efficacy in clinical trials even if the biomarker cannot replace clinical outcome as a surrogate. The building and evaluation of an association model between biomarkers and clinical outcomes are two equally important concerns regarding the prediction of clinical outcome. This paper is to address both issues in a Bayesian framework. A Bayesian meta-analytic approach is proposed to build a prediction model between the biomarker and clinical endpoint for dichotomous variables. Compared with other Bayesian methods, the proposed model only requires trial-level summary data of historical trials in model building. By using extensive simulations, we evaluate the link function and the application condition of the proposed Bayesian model under scenario (i) equal positive predictive value (PPV) and negative predictive value (NPV) and (ii) higher NPV and lower PPV. In the simulations, the patient-level data is generated to evaluate the meta-analytic model. PPV and NPV are employed to describe the patient-level relationship between the biomarker and the clinical outcome. The minimum number of historical trials to be included in building the model is also considered. It is seen from the simulations that the logit link function performs better than the odds and cloglog functions under both scenarios. PPV/NPV ≥0.5 for equal PPV and NPV, and PPV + NPV ≥1 for higher NPV and lower PPV are proposed in order to predict clinical outcome accurately and precisely when the proposed model is considered. Twenty historical trials are required to be included in model building when PPV and NPV are equal. For unequal PPV and NPV, the minimum number of historical trials for model building is proposed to be five. A hypothetical example shows an application of the proposed model in global drug development. The proposed Bayesian model is able to predict well the clinical endpoint from the observed biomarker data for dichotomous variables as long as the conditions are satisfied. It could be applied in drug development. But the practical problems in applications have to be studied in further research.

Learning Instance-Specific Predictive Models

PubMed Central

Visweswaran, Shyam; Cooper, Gregory F.

2013-01-01

This paper introduces a Bayesian algorithm for constructing predictive models from data that are optimized to predict a target variable well for a particular instance. This algorithm learns Markov blanket models, carries out Bayesian model averaging over a set of models to predict a target variable of the instance at hand, and employs an instance-specific heuristic to locate a set of suitable models to average over. We call this method the instance-specific Markov blanket (ISMB) algorithm. The ISMB algorithm was evaluated on 21 UCI data sets using five different performance measures and its performance was compared to that of several commonly used predictive algorithms, including nave Bayes, C4.5 decision tree, logistic regression, neural networks, k-Nearest Neighbor, Lazy Bayesian Rules, and AdaBoost. Over all the data sets, the ISMB algorithm performed better on average on all performance measures against all the comparison algorithms. PMID:25045325

Bayesian modeling of flexible cognitive control

PubMed Central

Jiang, Jiefeng; Heller, Katherine; Egner, Tobias

2014-01-01

“Cognitive control” describes endogenous guidance of behavior in situations where routine stimulus-response associations are suboptimal for achieving a desired goal. The computational and neural mechanisms underlying this capacity remain poorly understood. We examine recent advances stemming from the application of a Bayesian learner perspective that provides optimal prediction for control processes. In reviewing the application of Bayesian models to cognitive control, we note that an important limitation in current models is a lack of a plausible mechanism for the flexible adjustment of control over conflict levels changing at varying temporal scales. We then show that flexible cognitive control can be achieved by a Bayesian model with a volatility-driven learning mechanism that modulates dynamically the relative dependence on recent and remote experiences in its prediction of future control demand. We conclude that the emergent Bayesian perspective on computational mechanisms of cognitive control holds considerable promise, especially if future studies can identify neural substrates of the variables encoded by these models, and determine the nature (Bayesian or otherwise) of their neural implementation. PMID:24929218

Technical note: Bayesian calibration of dynamic ruminant nutrition models.

PubMed

Reed, K F; Arhonditsis, G B; France, J; Kebreab, E

2016-08-01

Mechanistic models of ruminant digestion and metabolism have advanced our understanding of the processes underlying ruminant animal physiology. Deterministic modeling practices ignore the inherent variation within and among individual animals and thus have no way to assess how sources of error influence model outputs. We introduce Bayesian calibration of mathematical models to address the need for robust mechanistic modeling tools that can accommodate error analysis by remaining within the bounds of data-based parameter estimation. For the purpose of prediction, the Bayesian approach generates a posterior predictive distribution that represents the current estimate of the value of the response variable, taking into account both the uncertainty about the parameters and model residual variability. Predictions are expressed as probability distributions, thereby conveying significantly more information than point estimates in regard to uncertainty. Our study illustrates some of the technical advantages of Bayesian calibration and discusses the future perspectives in the context of animal nutrition modeling. Copyright © 2016 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.

APPLICATION OF BAYESIAN MONTE CARLO ANALYSIS TO A LAGRANGIAN PHOTOCHEMICAL AIR QUALITY MODEL. (R824792)

EPA Science Inventory

Uncertainties in ozone concentrations predicted with a Lagrangian photochemical air quality model have been estimated using Bayesian Monte Carlo (BMC) analysis. Bayesian Monte Carlo analysis provides a means of combining subjective "prior" uncertainty estimates developed ...

Variational Bayesian identification and prediction of stochastic nonlinear dynamic causal models.

PubMed

Daunizeau, J; Friston, K J; Kiebel, S J

2009-11-01

In this paper, we describe a general variational Bayesian approach for approximate inference on nonlinear stochastic dynamic models. This scheme extends established approximate inference on hidden-states to cover: (i) nonlinear evolution and observation functions, (ii) unknown parameters and (precision) hyperparameters and (iii) model comparison and prediction under uncertainty. Model identification or inversion entails the estimation of the marginal likelihood or evidence of a model. This difficult integration problem can be finessed by optimising a free-energy bound on the evidence using results from variational calculus. This yields a deterministic update scheme that optimises an approximation to the posterior density on the unknown model variables. We derive such a variational Bayesian scheme in the context of nonlinear stochastic dynamic hierarchical models, for both model identification and time-series prediction. The computational complexity of the scheme is comparable to that of an extended Kalman filter, which is critical when inverting high dimensional models or long time-series. Using Monte-Carlo simulations, we assess the estimation efficiency of this variational Bayesian approach using three stochastic variants of chaotic dynamic systems. We also demonstrate the model comparison capabilities of the method, its self-consistency and its predictive power.

Bayesian Learning and the Psychology of Rule Induction

ERIC Educational Resources Information Center

Endress, Ansgar D.

2013-01-01

In recent years, Bayesian learning models have been applied to an increasing variety of domains. While such models have been criticized on theoretical grounds, the underlying assumptions and predictions are rarely made concrete and tested experimentally. Here, I use Frank and Tenenbaum's (2011) Bayesian model of rule-learning as a case study to…

Properties of the Bayesian Knowledge Tracing Model

ERIC Educational Resources Information Center

van de Sande, Brett

2013-01-01

Bayesian Knowledge Tracing is used very widely to model student learning. It comes in two different forms: The first form is the Bayesian Knowledge Tracing "hidden Markov model" which predicts the probability of correct application of a skill as a function of the number of previous opportunities to apply that skill and the model…

Evaluation of calibration efficacy under different levels of uncertainty

DOE PAGES

Heo, Yeonsook; Graziano, Diane J.; Guzowski, Leah; ...

2014-06-10

This study examines how calibration performs under different levels of uncertainty in model input data. It specifically assesses the efficacy of Bayesian calibration to enhance the reliability of EnergyPlus model predictions. A Bayesian approach can be used to update uncertain values of parameters, given measured energy-use data, and to quantify the associated uncertainty.We assess the efficacy of Bayesian calibration under a controlled virtual-reality setup, which enables rigorous validation of the accuracy of calibration results in terms of both calibrated parameter values and model predictions. Case studies demonstrate the performance of Bayesian calibration of base models developed from audit data withmore » differing levels of detail in building design, usage, and operation.« less

Bayesian regression models outperform partial least squares methods for predicting milk components and technological properties using infrared spectral data

PubMed Central

Ferragina, A.; de los Campos, G.; Vazquez, A. I.; Cecchinato, A.; Bittante, G.

2017-01-01

The aim of this study was to assess the performance of Bayesian models commonly used for genomic selection to predict “difficult-to-predict” dairy traits, such as milk fatty acid (FA) expressed as percentage of total fatty acids, and technological properties, such as fresh cheese yield and protein recovery, using Fourier-transform infrared (FTIR) spectral data. Our main hypothesis was that Bayesian models that can estimate shrinkage and perform variable selection may improve our ability to predict FA traits and technological traits above and beyond what can be achieved using the current calibration models (e.g., partial least squares, PLS). To this end, we assessed a series of Bayesian methods and compared their prediction performance with that of PLS. The comparison between models was done using the same sets of data (i.e., same samples, same variability, same spectral treatment) for each trait. Data consisted of 1,264 individual milk samples collected from Brown Swiss cows for which gas chromatographic FA composition, milk coagulation properties, and cheese-yield traits were available. For each sample, 2 spectra in the infrared region from 5,011 to 925 cm−1 were available and averaged before data analysis. Three Bayesian models: Bayesian ridge regression (Bayes RR), Bayes A, and Bayes B, and 2 reference models: PLS and modified PLS (MPLS) procedures, were used to calibrate equations for each of the traits. The Bayesian models used were implemented in the R package BGLR (http://cran.r-project.org/web/packages/BGLR/index.html), whereas the PLS and MPLS were those implemented in the WinISI II software (Infrasoft International LLC, State College, PA). Prediction accuracy was estimated for each trait and model using 25 replicates of a training-testing validation procedure. Compared with PLS, which is currently the most widely used calibration method, MPLS and the 3 Bayesian methods showed significantly greater prediction accuracy. Accuracy increased in moving from calibration to external validation methods, and in moving from PLS and MPLS to Bayesian methods, particularly Bayes A and Bayes B. The maximum R2 value of validation was obtained with Bayes B and Bayes A. For the FA, C10:0 (% of each FA on total FA basis) had the highest R2 (0.75, achieved with Bayes A and Bayes B), and among the technological traits, fresh cheese yield R2 of 0.82 (achieved with Bayes B). These 2 methods have proven to be useful instruments in shrinking and selecting very informative wavelengths and inferring the structure and functions of the analyzed traits. We conclude that Bayesian models are powerful tools for deriving calibration equations, and, importantly, these equations can be easily developed using existing open-source software. As part of our study, we provide scripts based on the open source R software BGLR, which can be used to train customized prediction equations for other traits or populations. PMID:26387015

Spatiotemporal Bayesian networks for malaria prediction.

PubMed

Haddawy, Peter; Hasan, A H M Imrul; Kasantikul, Rangwan; Lawpoolsri, Saranath; Sa-Angchai, Patiwat; Kaewkungwal, Jaranit; Singhasivanon, Pratap

2018-01-01

Targeted intervention and resource allocation are essential for effective malaria control, particularly in remote areas, with predictive models providing important information for decision making. While a diversity of modeling technique have been used to create predictive models of malaria, no work has made use of Bayesian networks. Bayes nets are attractive due to their ability to represent uncertainty, model time lagged and nonlinear relations, and provide explanations. This paper explores the use of Bayesian networks to model malaria, demonstrating the approach by creating village level models with weekly temporal resolution for Tha Song Yang district in northern Thailand. The networks are learned using data on cases and environmental covariates. Three types of networks are explored: networks for numeric prediction, networks for outbreak prediction, and networks that incorporate spatial autocorrelation. Evaluation of the numeric prediction network shows that the Bayes net has prediction accuracy in terms of mean absolute error of about 1.4 cases for 1 week prediction and 1.7 cases for 6 week prediction. The network for outbreak prediction has an ROC AUC above 0.9 for all prediction horizons. Comparison of prediction accuracy of both Bayes nets against several traditional modeling approaches shows the Bayes nets to outperform the other models for longer time horizon prediction of high incidence transmission. To model spread of malaria over space, we elaborate the models with links between the village networks. This results in some very large models which would be far too laborious to build by hand. So we represent the models as collections of probability logic rules and automatically generate the networks. Evaluation of the models shows that the autocorrelation links significantly improve prediction accuracy for some villages in regions of high incidence. We conclude that spatiotemporal Bayesian networks are a highly promising modeling alternative for prediction of malaria and other vector-borne diseases. Copyright © 2017 Elsevier B.V. All rights reserved.

A Bayesian Performance Prediction Model for Mathematics Education: A Prototypical Approach for Effective Group Composition

ERIC Educational Resources Information Center

Bekele, Rahel; McPherson, Maggie

2011-01-01

This research work presents a Bayesian Performance Prediction Model that was created in order to determine the strength of personality traits in predicting the level of mathematics performance of high school students in Addis Ababa. It is an automated tool that can be used to collect information from students for the purpose of effective group…

A Bayesian network model for predicting aquatic toxicity mode of action using two dimensional theoretical molecular descriptors

EPA Science Inventory

The mode of toxic action (MoA) has been recognized as a key determinant of chemical toxicity, but development of predictive MoA classification models in aquatic toxicology has been limited. We developed a Bayesian network model to classify aquatic toxicity MoA using a recently pu...

Predicting Graduation Rates at 4-Year Broad Access Institutions Using a Bayesian Modeling Approach

ERIC Educational Resources Information Center

Crisp, Gloria; Doran, Erin; Salis Reyes, Nicole A.

2018-01-01

This study models graduation rates at 4-year broad access institutions (BAIs). We examine the student body, structural-demographic, and financial characteristics that best predict 6-year graduation rates across two time periods (2008-2009 and 2014-2015). A Bayesian model averaging approach is utilized to account for uncertainty in variable…

Bayesian regression models outperform partial least squares methods for predicting milk components and technological properties using infrared spectral data.

PubMed

Ferragina, A; de los Campos, G; Vazquez, A I; Cecchinato, A; Bittante, G

2015-11-01

The aim of this study was to assess the performance of Bayesian models commonly used for genomic selection to predict "difficult-to-predict" dairy traits, such as milk fatty acid (FA) expressed as percentage of total fatty acids, and technological properties, such as fresh cheese yield and protein recovery, using Fourier-transform infrared (FTIR) spectral data. Our main hypothesis was that Bayesian models that can estimate shrinkage and perform variable selection may improve our ability to predict FA traits and technological traits above and beyond what can be achieved using the current calibration models (e.g., partial least squares, PLS). To this end, we assessed a series of Bayesian methods and compared their prediction performance with that of PLS. The comparison between models was done using the same sets of data (i.e., same samples, same variability, same spectral treatment) for each trait. Data consisted of 1,264 individual milk samples collected from Brown Swiss cows for which gas chromatographic FA composition, milk coagulation properties, and cheese-yield traits were available. For each sample, 2 spectra in the infrared region from 5,011 to 925 cm(-1) were available and averaged before data analysis. Three Bayesian models: Bayesian ridge regression (Bayes RR), Bayes A, and Bayes B, and 2 reference models: PLS and modified PLS (MPLS) procedures, were used to calibrate equations for each of the traits. The Bayesian models used were implemented in the R package BGLR (http://cran.r-project.org/web/packages/BGLR/index.html), whereas the PLS and MPLS were those implemented in the WinISI II software (Infrasoft International LLC, State College, PA). Prediction accuracy was estimated for each trait and model using 25 replicates of a training-testing validation procedure. Compared with PLS, which is currently the most widely used calibration method, MPLS and the 3 Bayesian methods showed significantly greater prediction accuracy. Accuracy increased in moving from calibration to external validation methods, and in moving from PLS and MPLS to Bayesian methods, particularly Bayes A and Bayes B. The maximum R(2) value of validation was obtained with Bayes B and Bayes A. For the FA, C10:0 (% of each FA on total FA basis) had the highest R(2) (0.75, achieved with Bayes A and Bayes B), and among the technological traits, fresh cheese yield R(2) of 0.82 (achieved with Bayes B). These 2 methods have proven to be useful instruments in shrinking and selecting very informative wavelengths and inferring the structure and functions of the analyzed traits. We conclude that Bayesian models are powerful tools for deriving calibration equations, and, importantly, these equations can be easily developed using existing open-source software. As part of our study, we provide scripts based on the open source R software BGLR, which can be used to train customized prediction equations for other traits or populations. Copyright © 2015 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.

Bayesian Fundamentalism or Enlightenment? On the explanatory status and theoretical contributions of Bayesian models of cognition.

PubMed

Jones, Matt; Love, Bradley C

2011-08-01

The prominence of Bayesian modeling of cognition has increased recently largely because of mathematical advances in specifying and deriving predictions from complex probabilistic models. Much of this research aims to demonstrate that cognitive behavior can be explained from rational principles alone, without recourse to psychological or neurological processes and representations. We note commonalities between this rational approach and other movements in psychology - namely, Behaviorism and evolutionary psychology - that set aside mechanistic explanations or make use of optimality assumptions. Through these comparisons, we identify a number of challenges that limit the rational program's potential contribution to psychological theory. Specifically, rational Bayesian models are significantly unconstrained, both because they are uninformed by a wide range of process-level data and because their assumptions about the environment are generally not grounded in empirical measurement. The psychological implications of most Bayesian models are also unclear. Bayesian inference itself is conceptually trivial, but strong assumptions are often embedded in the hypothesis sets and the approximation algorithms used to derive model predictions, without a clear delineation between psychological commitments and implementational details. Comparing multiple Bayesian models of the same task is rare, as is the realization that many Bayesian models recapitulate existing (mechanistic level) theories. Despite the expressive power of current Bayesian models, we argue they must be developed in conjunction with mechanistic considerations to offer substantive explanations of cognition. We lay out several means for such an integration, which take into account the representations on which Bayesian inference operates, as well as the algorithms and heuristics that carry it out. We argue this unification will better facilitate lasting contributions to psychological theory, avoiding the pitfalls that have plagued previous theoretical movements.

Estimating Tree Height-Diameter Models with the Bayesian Method

PubMed Central

Duan, Aiguo; Zhang, Jianguo; Xiang, Congwei

2014-01-01

Six candidate height-diameter models were used to analyze the height-diameter relationships. The common methods for estimating the height-diameter models have taken the classical (frequentist) approach based on the frequency interpretation of probability, for example, the nonlinear least squares method (NLS) and the maximum likelihood method (ML). The Bayesian method has an exclusive advantage compared with classical method that the parameters to be estimated are regarded as random variables. In this study, the classical and Bayesian methods were used to estimate six height-diameter models, respectively. Both the classical method and Bayesian method showed that the Weibull model was the “best” model using data1. In addition, based on the Weibull model, data2 was used for comparing Bayesian method with informative priors with uninformative priors and classical method. The results showed that the improvement in prediction accuracy with Bayesian method led to narrower confidence bands of predicted value in comparison to that for the classical method, and the credible bands of parameters with informative priors were also narrower than uninformative priors and classical method. The estimated posterior distributions for parameters can be set as new priors in estimating the parameters using data2. PMID:24711733

Estimating tree height-diameter models with the Bayesian method.

PubMed

Zhang, Xiongqing; Duan, Aiguo; Zhang, Jianguo; Xiang, Congwei

2014-01-01

Six candidate height-diameter models were used to analyze the height-diameter relationships. The common methods for estimating the height-diameter models have taken the classical (frequentist) approach based on the frequency interpretation of probability, for example, the nonlinear least squares method (NLS) and the maximum likelihood method (ML). The Bayesian method has an exclusive advantage compared with classical method that the parameters to be estimated are regarded as random variables. In this study, the classical and Bayesian methods were used to estimate six height-diameter models, respectively. Both the classical method and Bayesian method showed that the Weibull model was the "best" model using data1. In addition, based on the Weibull model, data2 was used for comparing Bayesian method with informative priors with uninformative priors and classical method. The results showed that the improvement in prediction accuracy with Bayesian method led to narrower confidence bands of predicted value in comparison to that for the classical method, and the credible bands of parameters with informative priors were also narrower than uninformative priors and classical method. The estimated posterior distributions for parameters can be set as new priors in estimating the parameters using data2.

Genomic-Enabled Prediction of Ordinal Data with Bayesian Logistic Ordinal Regression.

PubMed

Montesinos-López, Osval A; Montesinos-López, Abelardo; Crossa, José; Burgueño, Juan; Eskridge, Kent

2015-08-18

Most genomic-enabled prediction models developed so far assume that the response variable is continuous and normally distributed. The exception is the probit model, developed for ordered categorical phenotypes. In statistical applications, because of the easy implementation of the Bayesian probit ordinal regression (BPOR) model, Bayesian logistic ordinal regression (BLOR) is implemented rarely in the context of genomic-enabled prediction [sample size (n) is much smaller than the number of parameters (p)]. For this reason, in this paper we propose a BLOR model using the Pólya-Gamma data augmentation approach that produces a Gibbs sampler with similar full conditional distributions of the BPOR model and with the advantage that the BPOR model is a particular case of the BLOR model. We evaluated the proposed model by using simulation and two real data sets. Results indicate that our BLOR model is a good alternative for analyzing ordinal data in the context of genomic-enabled prediction with the probit or logit link. Copyright © 2015 Montesinos-López et al.

Using Bayesian Networks for Candidate Generation in Consistency-based Diagnosis

NASA Technical Reports Server (NTRS)

Narasimhan, Sriram; Mengshoel, Ole

2008-01-01

Consistency-based diagnosis relies heavily on the assumption that discrepancies between model predictions and sensor observations can be detected accurately. When sources of uncertainty like sensor noise and model abstraction exist robust schemes have to be designed to make a binary decision on whether predictions are consistent with observations. This risks the occurrence of false alarms and missed alarms when an erroneous decision is made. Moreover when multiple sensors (with differing sensing properties) are available the degree of match between predictions and observations can be used to guide the search for fault candidates. In this paper we propose a novel approach to handle this problem using Bayesian networks. In the consistency- based diagnosis formulation, automatically generated Bayesian networks are used to encode a probabilistic measure of fit between predictions and observations. A Bayesian network inference algorithm is used to compute most probable fault candidates.

A Fast Surrogate-facilitated Data-driven Bayesian Approach to Uncertainty Quantification of a Regional Groundwater Flow Model with Structural Error

NASA Astrophysics Data System (ADS)

Xu, T.; Valocchi, A. J.; Ye, M.; Liang, F.

2016-12-01

Due to simplification and/or misrepresentation of the real aquifer system, numerical groundwater flow and solute transport models are usually subject to model structural error. During model calibration, the hydrogeological parameters may be overly adjusted to compensate for unknown structural error. This may result in biased predictions when models are used to forecast aquifer response to new forcing. In this study, we extend a fully Bayesian method [Xu and Valocchi, 2015] to calibrate a real-world, regional groundwater flow model. The method uses a data-driven error model to describe model structural error and jointly infers model parameters and structural error. In this study, Bayesian inference is facilitated using high performance computing and fast surrogate models. The surrogate models are constructed using machine learning techniques to emulate the response simulated by the computationally expensive groundwater model. We demonstrate in the real-world case study that explicitly accounting for model structural error yields parameter posterior distributions that are substantially different from those derived by the classical Bayesian calibration that does not account for model structural error. In addition, the Bayesian with error model method gives significantly more accurate prediction along with reasonable credible intervals.

An introduction to using Bayesian linear regression with clinical data.

PubMed

Baldwin, Scott A; Larson, Michael J

2017-11-01

Statistical training psychology focuses on frequentist methods. Bayesian methods are an alternative to standard frequentist methods. This article provides researchers with an introduction to fundamental ideas in Bayesian modeling. We use data from an electroencephalogram (EEG) and anxiety study to illustrate Bayesian models. Specifically, the models examine the relationship between error-related negativity (ERN), a particular event-related potential, and trait anxiety. Methodological topics covered include: how to set up a regression model in a Bayesian framework, specifying priors, examining convergence of the model, visualizing and interpreting posterior distributions, interval estimates, expected and predicted values, and model comparison tools. We also discuss situations where Bayesian methods can outperform frequentist methods as well has how to specify more complicated regression models. Finally, we conclude with recommendations about reporting guidelines for those using Bayesian methods in their own research. We provide data and R code for replicating our analyses. Copyright © 2017 Elsevier Ltd. All rights reserved.

Predicting Student Success: A Naïve Bayesian Application to Community College Data

ERIC Educational Resources Information Center

Ornelas, Fermin; Ordonez, Carlos

2017-01-01

This research focuses on developing and implementing a continuous Naïve Bayesian classifier for GEAR courses at Rio Salado Community College. Previous implementation efforts of a discrete version did not predict as well, 70%, and had deployment issues. This predictive model has higher prediction, over 90%, accuracy for both at-risk and successful…

Predicting ICU mortality: a comparison of stationary and nonstationary temporal models.

PubMed Central

Kayaalp, M.; Cooper, G. F.; Clermont, G.

2000-01-01

OBJECTIVE: This study evaluates the effectiveness of the stationarity assumption in predicting the mortality of intensive care unit (ICU) patients at the ICU discharge. DESIGN: This is a comparative study. A stationary temporal Bayesian network learned from data was compared to a set of (33) nonstationary temporal Bayesian networks learned from data. A process observed as a sequence of events is stationary if its stochastic properties stay the same when the sequence is shifted in a positive or negative direction by a constant time parameter. The temporal Bayesian networks forecast mortalities of patients, where each patient has one record per day. The predictive performance of the stationary model is compared with nonstationary models using the area under the receiver operating characteristics (ROC) curves. RESULTS: The stationary model usually performed best. However, one nonstationary model using large data sets performed significantly better than the stationary model. CONCLUSION: Results suggest that using a combination of stationary and nonstationary models may predict better than using either alone. PMID:11079917

Bayesian network modeling applied to coastal geomorphology: lessons learned from a decade of experimentation and application

NASA Astrophysics Data System (ADS)

Plant, N. G.; Thieler, E. R.; Gutierrez, B.; Lentz, E. E.; Zeigler, S. L.; Van Dongeren, A.; Fienen, M. N.

2016-12-01

We evaluate the strengths and weaknesses of Bayesian networks that have been used to address scientific and decision-support questions related to coastal geomorphology. We will provide an overview of coastal geomorphology research that has used Bayesian networks and describe what this approach can do and when it works (or fails to work). Over the past decade, Bayesian networks have been formulated to analyze the multi-variate structure and evolution of coastal morphology and associated human and ecological impacts. The approach relates observable system variables to each other by estimating discrete correlations. The resulting Bayesian-networks make predictions that propagate errors, conduct inference via Bayes rule, or both. In scientific applications, the model results are useful for hypothesis testing, using confidence estimates to gage the strength of tests while applications to coastal resource management are aimed at decision-support, where the probabilities of desired ecosystems outcomes are evaluated. The range of Bayesian-network applications to coastal morphology includes emulation of high-resolution wave transformation models to make oceanographic predictions, morphologic response to storms and/or sea-level rise, groundwater response to sea-level rise and morphologic variability, habitat suitability for endangered species, and assessment of monetary or human-life risk associated with storms. All of these examples are based on vast observational data sets, numerical model output, or both. We will discuss the progression of our experiments, which has included testing whether the Bayesian-network approach can be implemented and is appropriate for addressing basic and applied scientific problems and evaluating the hindcast and forecast skill of these implementations. We will present and discuss calibration/validation tests that are used to assess the robustness of Bayesian-network models and we will compare these results to tests of other models. This will demonstrate how Bayesian networks are used to extract new insights about coastal morphologic behavior, assess impacts to societal and ecological systems, and communicate probabilistic predictions to decision makers.

Prediction of Individual Serum Infliximab Concentrations in Inflammatory Bowel Disease by a Bayesian Dashboard System.

PubMed

Eser, Alexander; Primas, Christian; Reinisch, Sieglinde; Vogelsang, Harald; Novacek, Gottfried; Mould, Diane R; Reinisch, Walter

2018-01-30

Despite a robust exposure-response relationship of infliximab in inflammatory bowel disease (IBD), attempts to adjust dosing to individually predicted serum concentrations of infliximab (SICs) are lacking. Compared with labor-intensive conventional software for pharmacokinetic (PK) modeling (eg, NONMEM) dashboards are easy-to-use programs incorporating complex Bayesian statistics to determine individual pharmacokinetics. We evaluated various infliximab detection assays and the number of samples needed to precisely forecast individual SICs using a Bayesian dashboard. We assessed long-term infliximab retention in patients being dosed concordantly versus discordantly with Bayesian dashboard recommendations. Three hundred eighty-two serum samples from 117 adult IBD patients on infliximab maintenance therapy were analyzed by 3 commercially available assays. Data from each assay was modeled using NONMEM and a Bayesian dashboard. PK parameter precision and residual variability were assessed. Forecast concentrations from both systems were compared with observed concentrations. Infliximab retention was assessed by prediction for dose intensification via Bayesian dashboard versus real-life practice. Forecast precision of SICs varied between detection assays. At least 3 SICs from a reliable assay are needed for an accurate forecast. The Bayesian dashboard performed similarly to NONMEM to predict SICs. Patients dosed concordantly with Bayesian dashboard recommendations had a significantly longer median drug survival than those dosed discordantly (51.5 versus 4.6 months, P < .0001). The Bayesian dashboard helps to assess the diagnostic performance of infliximab detection assays. Three, not single, SICs provide sufficient information for individualized dose adjustment when incorporated into the Bayesian dashboard. Treatment adjusted to forecasted SICs is associated with longer drug retention of infliximab. © 2018, The American College of Clinical Pharmacology.

Bayesian model reduction and empirical Bayes for group (DCM) studies

PubMed Central

Friston, Karl J.; Litvak, Vladimir; Oswal, Ashwini; Razi, Adeel; Stephan, Klaas E.; van Wijk, Bernadette C.M.; Ziegler, Gabriel; Zeidman, Peter

2016-01-01

This technical note describes some Bayesian procedures for the analysis of group studies that use nonlinear models at the first (within-subject) level – e.g., dynamic causal models – and linear models at subsequent (between-subject) levels. Its focus is on using Bayesian model reduction to finesse the inversion of multiple models of a single dataset or a single (hierarchical or empirical Bayes) model of multiple datasets. These applications of Bayesian model reduction allow one to consider parametric random effects and make inferences about group effects very efficiently (in a few seconds). We provide the relatively straightforward theoretical background to these procedures and illustrate their application using a worked example. This example uses a simulated mismatch negativity study of schizophrenia. We illustrate the robustness of Bayesian model reduction to violations of the (commonly used) Laplace assumption in dynamic causal modelling and show how its recursive application can facilitate both classical and Bayesian inference about group differences. Finally, we consider the application of these empirical Bayesian procedures to classification and prediction. PMID:26569570

Predicting site locations for biomass using facilities with Bayesian methods

Treesearch

Timothy M. Young; James H. Perdue; Xia Huang

2017-01-01

Logistic regression models combined with Bayesian inference were developed to predict locations and quantify factors that influence the siting of biomass-using facilities that use woody biomass in the Southeastern United States. Predictions were developed for two groups of mills, one representing larger capacity mills similar to pulp and paper mills (Group II...

Bayesian Forecasting Tool to Predict the Need for Antidote in Acute Acetaminophen Overdose.

PubMed

Desrochers, Julie; Wojciechowski, Jessica; Klein-Schwartz, Wendy; Gobburu, Jogarao V S; Gopalakrishnan, Mathangi

2017-08-01

Acetaminophen (APAP) overdose is the leading cause of acute liver injury in the United States. Patients with elevated plasma acetaminophen concentrations (PACs) require hepatoprotective treatment with N-acetylcysteine (NAC). These patients have been primarily risk-stratified using the Rumack-Matthew nomogram. Previous studies of acute APAP overdoses found that the nomogram failed to accurately predict the need for the antidote. The objectives of this study were to develop a population pharmacokinetic (PK) model for APAP following acute overdose and evaluate the utility of population PK model-based Bayesian forecasting in NAC administration decisions. Limited APAP concentrations from a retrospective cohort of acute overdosed subjects from the Maryland Poison Center were used to develop the population PK model and to investigate the effect of type of APAP products and other prognostic factors. The externally validated population PK model was used a prior for Bayesian forecasting to predict the individual PK profile when one or two observed PACs were available. The utility of Bayesian forecasted APAP concentration-time profiles inferred from one (first) or two (first and second) PAC observations were also tested in their ability to predict the observed NAC decisions. A one-compartment model with first-order absorption and elimination adequately described the data with single activated charcoal and APAP products as significant covariates on absorption and bioavailability. The Bayesian forecasted individual concentration-time profiles had acceptable bias (6.2% and 9.8%) and accuracy (40.5% and 41.9%) when either one or two PACs were considered, respectively. The sensitivity and negative predictive value of the Bayesian forecasted NAC decisions using one PAC were 84% and 92.6%, respectively. The population PK analysis provided a platform for acceptably predicting an individual's concentration-time profile following acute APAP overdose with at least one PAC, and the individual's covariate profile, and can potentially be used for making early NAC administration decisions. © 2017 Pharmacotherapy Publications, Inc.

Assessment of parametric uncertainty for groundwater reactive transport modeling,

USGS Publications Warehouse

Shi, Xiaoqing; Ye, Ming; Curtis, Gary P.; Miller, Geoffery L.; Meyer, Philip D.; Kohler, Matthias; Yabusaki, Steve; Wu, Jichun

2014-01-01

The validity of using Gaussian assumptions for model residuals in uncertainty quantification of a groundwater reactive transport model was evaluated in this study. Least squares regression methods explicitly assume Gaussian residuals, and the assumption leads to Gaussian likelihood functions, model parameters, and model predictions. While the Bayesian methods do not explicitly require the Gaussian assumption, Gaussian residuals are widely used. This paper shows that the residuals of the reactive transport model are non-Gaussian, heteroscedastic, and correlated in time; characterizing them requires using a generalized likelihood function such as the formal generalized likelihood function developed by Schoups and Vrugt (2010). For the surface complexation model considered in this study for simulating uranium reactive transport in groundwater, parametric uncertainty is quantified using the least squares regression methods and Bayesian methods with both Gaussian and formal generalized likelihood functions. While the least squares methods and Bayesian methods with Gaussian likelihood function produce similar Gaussian parameter distributions, the parameter distributions of Bayesian uncertainty quantification using the formal generalized likelihood function are non-Gaussian. In addition, predictive performance of formal generalized likelihood function is superior to that of least squares regression and Bayesian methods with Gaussian likelihood function. The Bayesian uncertainty quantification is conducted using the differential evolution adaptive metropolis (DREAM(zs)) algorithm; as a Markov chain Monte Carlo (MCMC) method, it is a robust tool for quantifying uncertainty in groundwater reactive transport models. For the surface complexation model, the regression-based local sensitivity analysis and Morris- and DREAM(ZS)-based global sensitivity analysis yield almost identical ranking of parameter importance. The uncertainty analysis may help select appropriate likelihood functions, improve model calibration, and reduce predictive uncertainty in other groundwater reactive transport and environmental modeling.

Bayesian model checking: A comparison of tests

NASA Astrophysics Data System (ADS)

Lucy, L. B.

2018-06-01

Two procedures for checking Bayesian models are compared using a simple test problem based on the local Hubble expansion. Over four orders of magnitude, p-values derived from a global goodness-of-fit criterion for posterior probability density functions agree closely with posterior predictive p-values. The former can therefore serve as an effective proxy for the difficult-to-calculate posterior predictive p-values.

Prediction and assimilation of surf-zone processes using a Bayesian network: Part II: Inverse models

USGS Publications Warehouse

Plant, Nathaniel G.; Holland, K. Todd

2011-01-01

A Bayesian network model has been developed to simulate a relatively simple problem of wave propagation in the surf zone (detailed in Part I). Here, we demonstrate that this Bayesian model can provide both inverse modeling and data-assimilation solutions for predicting offshore wave heights and depth estimates given limited wave-height and depth information from an onshore location. The inverse method is extended to allow data assimilation using observational inputs that are not compatible with deterministic solutions of the problem. These inputs include sand bar positions (instead of bathymetry) and estimates of the intensity of wave breaking (instead of wave-height observations). Our results indicate that wave breaking information is essential to reduce prediction errors. In many practical situations, this information could be provided from a shore-based observer or from remote-sensing systems. We show that various combinations of the assimilated inputs significantly reduce the uncertainty in the estimates of water depths and wave heights in the model domain. Application of the Bayesian network model to new field data demonstrated significant predictive skill (R2 = 0.7) for the inverse estimate of a month-long time series of offshore wave heights. The Bayesian inverse results include uncertainty estimates that were shown to be most accurate when given uncertainty in the inputs (e.g., depth and tuning parameters). Furthermore, the inverse modeling was extended to directly estimate tuning parameters associated with the underlying wave-process model. The inverse estimates of the model parameters not only showed an offshore wave height dependence consistent with results of previous studies but the uncertainty estimates of the tuning parameters also explain previously reported variations in the model parameters.

A study of finite mixture model: Bayesian approach on financial time series data

NASA Astrophysics Data System (ADS)

Phoong, Seuk-Yen; Ismail, Mohd Tahir

2014-07-01

Recently, statistician have emphasized on the fitting finite mixture model by using Bayesian method. Finite mixture model is a mixture of distributions in modeling a statistical distribution meanwhile Bayesian method is a statistical method that use to fit the mixture model. Bayesian method is being used widely because it has asymptotic properties which provide remarkable result. In addition, Bayesian method also shows consistency characteristic which means the parameter estimates are close to the predictive distributions. In the present paper, the number of components for mixture model is studied by using Bayesian Information Criterion. Identify the number of component is important because it may lead to an invalid result. Later, the Bayesian method is utilized to fit the k-component mixture model in order to explore the relationship between rubber price and stock market price for Malaysia, Thailand, Philippines and Indonesia. Lastly, the results showed that there is a negative effect among rubber price and stock market price for all selected countries.

A Bayesian Approach to Person Fit Analysis in Item Response Theory Models. Research Report.

ERIC Educational Resources Information Center

Glas, Cees A. W.; Meijer, Rob R.

A Bayesian approach to the evaluation of person fit in item response theory (IRT) models is presented. In a posterior predictive check, the observed value on a discrepancy variable is positioned in its posterior distribution. In a Bayesian framework, a Markov Chain Monte Carlo procedure can be used to generate samples of the posterior distribution…

Prediction of community prevalence of human onchocerciasis in the Amazonian onchocerciasis focus: Bayesian approach.

PubMed Central

Carabin, Hélène; Escalona, Marisela; Marshall, Clare; Vivas-Martínez, Sarai; Botto, Carlos; Joseph, Lawrence; Basáñez, María-Gloria

2003-01-01

OBJECTIVE: To develop a Bayesian hierarchical model for human onchocerciasis with which to explore the factors that influence prevalence of microfilariae in the Amazonian focus of onchocerciasis and predict the probability of any community being at least mesoendemic (>20% prevalence of microfilariae), and thus in need of priority ivermectin treatment. METHODS: Models were developed with data from 732 individuals aged > or =15 years who lived in 29 Yanomami communities along four rivers of the south Venezuelan Orinoco basin. The models' abilities to predict prevalences of microfilariae in communities were compared. The deviance information criterion, Bayesian P-values, and residual values were used to select the best model with an approximate cross-validation procedure. FINDINGS: A three-level model that acknowledged clustering of infection within communities performed best, with host age and sex included at the individual level, a river-dependent altitude effect at the community level, and additional clustering of communities along rivers. This model correctly classified 25/29 (86%) villages with respect to their need for priority ivermectin treatment. CONCLUSION: Bayesian methods are a flexible and useful approach for public health research and control planning. Our model acknowledges the clustering of infection within communities, allows investigation of links between individual- or community-specific characteristics and infection, incorporates additional uncertainty due to missing covariate data, and informs policy decisions by predicting the probability that a new community is at least mesoendemic. PMID:12973640

Prediction of community prevalence of human onchocerciasis in the Amazonian onchocerciasis focus: Bayesian approach.

PubMed

Carabin, Hélène; Escalona, Marisela; Marshall, Clare; Vivas-Martínez, Sarai; Botto, Carlos; Joseph, Lawrence; Basáñez, María-Gloria

2003-01-01

To develop a Bayesian hierarchical model for human onchocerciasis with which to explore the factors that influence prevalence of microfilariae in the Amazonian focus of onchocerciasis and predict the probability of any community being at least mesoendemic (>20% prevalence of microfilariae), and thus in need of priority ivermectin treatment. Models were developed with data from 732 individuals aged > or =15 years who lived in 29 Yanomami communities along four rivers of the south Venezuelan Orinoco basin. The models' abilities to predict prevalences of microfilariae in communities were compared. The deviance information criterion, Bayesian P-values, and residual values were used to select the best model with an approximate cross-validation procedure. A three-level model that acknowledged clustering of infection within communities performed best, with host age and sex included at the individual level, a river-dependent altitude effect at the community level, and additional clustering of communities along rivers. This model correctly classified 25/29 (86%) villages with respect to their need for priority ivermectin treatment. Bayesian methods are a flexible and useful approach for public health research and control planning. Our model acknowledges the clustering of infection within communities, allows investigation of links between individual- or community-specific characteristics and infection, incorporates additional uncertainty due to missing covariate data, and informs policy decisions by predicting the probability that a new community is at least mesoendemic.

Artificial and Bayesian Neural Networks

PubMed

Korhani Kangi, Azam; Bahrampour, Abbas

2018-02-26

Introduction and purpose: In recent years the use of neural networks without any premises for investigation of prognosis in analyzing survival data has increased. Artificial neural networks (ANN) use small processors with a continuous network to solve problems inspired by the human brain. Bayesian neural networks (BNN) constitute a neural-based approach to modeling and non-linearization of complex issues using special algorithms and statistical methods. Gastric cancer incidence is the first and third ranking for men and women in Iran, respectively. The aim of the present study was to assess the value of an artificial neural network and a Bayesian neural network for modeling and predicting of probability of gastric cancer patient death. Materials and Methods: In this study, we used information on 339 patients aged from 20 to 90 years old with positive gastric cancer, referred to Afzalipoor and Shahid Bahonar Hospitals in Kerman City from 2001 to 2015. The three layers perceptron neural network (ANN) and the Bayesian neural network (BNN) were used for predicting the probability of mortality using the available data. To investigate differences between the models, sensitivity, specificity, accuracy and the area under receiver operating characteristic curves (AUROCs) were generated. Results: In this study, the sensitivity and specificity of the artificial neural network and Bayesian neural network models were 0.882, 0.903 and 0.954, 0.909, respectively. Prediction accuracy and the area under curve ROC for the two models were 0.891, 0.944 and 0.935, 0.961. The age at diagnosis of gastric cancer was most important for predicting survival, followed by tumor grade, morphology, gender, smoking history, opium consumption, receiving chemotherapy, presence of metastasis, tumor stage, receiving radiotherapy, and being resident in a village. Conclusion: The findings of the present study indicated that the Bayesian neural network is preferable to an artificial neural network for predicting survival of gastric cancer patients in Iran. Creative Commons Attribution License

Bayesian model reduction and empirical Bayes for group (DCM) studies.

PubMed

Friston, Karl J; Litvak, Vladimir; Oswal, Ashwini; Razi, Adeel; Stephan, Klaas E; van Wijk, Bernadette C M; Ziegler, Gabriel; Zeidman, Peter

2016-03-01

This technical note describes some Bayesian procedures for the analysis of group studies that use nonlinear models at the first (within-subject) level - e.g., dynamic causal models - and linear models at subsequent (between-subject) levels. Its focus is on using Bayesian model reduction to finesse the inversion of multiple models of a single dataset or a single (hierarchical or empirical Bayes) model of multiple datasets. These applications of Bayesian model reduction allow one to consider parametric random effects and make inferences about group effects very efficiently (in a few seconds). We provide the relatively straightforward theoretical background to these procedures and illustrate their application using a worked example. This example uses a simulated mismatch negativity study of schizophrenia. We illustrate the robustness of Bayesian model reduction to violations of the (commonly used) Laplace assumption in dynamic causal modelling and show how its recursive application can facilitate both classical and Bayesian inference about group differences. Finally, we consider the application of these empirical Bayesian procedures to classification and prediction. Copyright © 2015 The Authors. Published by Elsevier Inc. All rights reserved.

Modelling the Flow Stress of Alloy 316L using a Multi-Layered Feed Forward Neural Network with Bayesian Regularization

NASA Astrophysics Data System (ADS)

Abiriand Bhekisipho Twala, Olufunminiyi

2017-08-01

In this paper, a multilayer feedforward neural network with Bayesian regularization constitutive model is developed for alloy 316L during high strain rate and high temperature plastic deformation. The input variables are strain rate, temperature and strain while the output value is the flow stress of the material. The results show that the use of Bayesian regularized technique reduces the potential of overfitting and overtraining. The prediction quality of the model is thereby improved. The model predictions are in good agreement with experimental measurements. The measurement data used for the network training and model comparison were taken from relevant literature. The developed model is robust as it can be generalized to deformation conditions slightly below or above the training dataset.

Dynamic Bayesian Networks for Student Modeling

ERIC Educational Resources Information Center

Kaser, Tanja; Klingler, Severin; Schwing, Alexander G.; Gross, Markus

2017-01-01

Intelligent tutoring systems adapt the curriculum to the needs of the individual student. Therefore, an accurate representation and prediction of student knowledge is essential. Bayesian Knowledge Tracing (BKT) is a popular approach for student modeling. The structure of BKT models, however, makes it impossible to represent the hierarchy and…

On the predictive information criteria for model determination in seismic hazard analysis

NASA Astrophysics Data System (ADS)

Varini, Elisa; Rotondi, Renata

2016-04-01

Many statistical tools have been developed for evaluating, understanding, and comparing models, from both frequentist and Bayesian perspectives. In particular, the problem of model selection can be addressed according to whether the primary goal is explanation or, alternatively, prediction. In the former case, the criteria for model selection are defined over the parameter space whose physical interpretation can be difficult; in the latter case, they are defined over the space of the observations, which has a more direct physical meaning. In the frequentist approaches, model selection is generally based on an asymptotic approximation which may be poor for small data sets (e.g. the F-test, the Kolmogorov-Smirnov test, etc.); moreover, these methods often apply under specific assumptions on models (e.g. models have to be nested in the likelihood ratio test). In the Bayesian context, among the criteria for explanation, the ratio of the observed marginal densities for two competing models, named Bayes Factor (BF), is commonly used for both model choice and model averaging (Kass and Raftery, J. Am. Stat. Ass., 1995). But BF does not apply to improper priors and, even when the prior is proper, it is not robust to the specification of the prior. These limitations can be extended to two famous penalized likelihood methods as the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC), since they are proved to be approximations of -2log BF . In the perspective that a model is as good as its predictions, the predictive information criteria aim at evaluating the predictive accuracy of Bayesian models or, in other words, at estimating expected out-of-sample prediction error using a bias-correction adjustment of within-sample error (Gelman et al., Stat. Comput., 2014). In particular, the Watanabe criterion is fully Bayesian because it averages the predictive distribution over the posterior distribution of parameters rather than conditioning on a point estimate, but it is hardly applicable to data which are not independent given parameters (Watanabe, J. Mach. Learn. Res., 2010). A solution is given by Ando and Tsay criterion where the joint density may be decomposed into the product of the conditional densities (Ando and Tsay, Int. J. Forecast., 2010). The above mentioned criteria are global summary measures of model performance, but more detailed analysis could be required to discover the reasons for poor global performance. In this latter case, a retrospective predictive analysis is performed on each individual observation. In this study we performed the Bayesian analysis of Italian data sets by four versions of a long-term hazard model known as the stress release model (Vere-Jones, J. Physics Earth, 1978; Bebbington and Harte, Geophys. J. Int., 2003; Varini and Rotondi, Environ. Ecol. Stat., 2015). Then we illustrate the results on their performance evaluated by Bayes Factor, predictive information criteria and retrospective predictive analysis.

Variations on Bayesian Prediction and Inference

DTIC Science & Technology

2016-05-09

inference 2.2.1 Background There are a number of statistical inference problems that are not generally formulated via a full probability model...problem of inference about an unknown parameter, the Bayesian approach requires a full probability 1. REPORT DATE (DD-MM-YYYY) 4. TITLE AND...the problem of inference about an unknown parameter, the Bayesian approach requires a full probability model/likelihood which can be an obstacle

A Bayesian approach to model structural error and input variability in groundwater modeling

NASA Astrophysics Data System (ADS)

Xu, T.; Valocchi, A. J.; Lin, Y. F. F.; Liang, F.

2015-12-01

Effective water resource management typically relies on numerical models to analyze groundwater flow and solute transport processes. Model structural error (due to simplification and/or misrepresentation of the "true" environmental system) and input forcing variability (which commonly arises since some inputs are uncontrolled or estimated with high uncertainty) are ubiquitous in groundwater models. Calibration that overlooks errors in model structure and input data can lead to biased parameter estimates and compromised predictions. We present a fully Bayesian approach for a complete assessment of uncertainty for spatially distributed groundwater models. The approach explicitly recognizes stochastic input and uses data-driven error models based on nonparametric kernel methods to account for model structural error. We employ exploratory data analysis to assist in specifying informative prior for error models to improve identifiability. The inference is facilitated by an efficient sampling algorithm based on DREAM-ZS and a parameter subspace multiple-try strategy to reduce the required number of forward simulations of the groundwater model. We demonstrate the Bayesian approach through a synthetic case study of surface-ground water interaction under changing pumping conditions. It is found that explicit treatment of errors in model structure and input data (groundwater pumping rate) has substantial impact on the posterior distribution of groundwater model parameters. Using error models reduces predictive bias caused by parameter compensation. In addition, input variability increases parametric and predictive uncertainty. The Bayesian approach allows for a comparison among the contributions from various error sources, which could inform future model improvement and data collection efforts on how to best direct resources towards reducing predictive uncertainty.

Bayesian Unimodal Density Regression for Causal Inference

ERIC Educational Resources Information Center

Karabatsos, George; Walker, Stephen G.

2011-01-01

Karabatsos and Walker (2011) introduced a new Bayesian nonparametric (BNP) regression model. Through analyses of real and simulated data, they showed that the BNP regression model outperforms other parametric and nonparametric regression models of common use, in terms of predictive accuracy of the outcome (dependent) variable. The other,…

Bayesian statistics in medicine: a 25 year review.

PubMed

Ashby, Deborah

2006-11-15

This review examines the state of Bayesian thinking as Statistics in Medicine was launched in 1982, reflecting particularly on its applicability and uses in medical research. It then looks at each subsequent five-year epoch, with a focus on papers appearing in Statistics in Medicine, putting these in the context of major developments in Bayesian thinking and computation with reference to important books, landmark meetings and seminal papers. It charts the growth of Bayesian statistics as it is applied to medicine and makes predictions for the future. From sparse beginnings, where Bayesian statistics was barely mentioned, Bayesian statistics has now permeated all the major areas of medical statistics, including clinical trials, epidemiology, meta-analyses and evidence synthesis, spatial modelling, longitudinal modelling, survival modelling, molecular genetics and decision-making in respect of new technologies.

CGBayesNets: Conditional Gaussian Bayesian Network Learning and Inference with Mixed Discrete and Continuous Data

PubMed Central

Weiss, Scott T.

2014-01-01

Bayesian Networks (BN) have been a popular predictive modeling formalism in bioinformatics, but their application in modern genomics has been slowed by an inability to cleanly handle domains with mixed discrete and continuous variables. Existing free BN software packages either discretize continuous variables, which can lead to information loss, or do not include inference routines, which makes prediction with the BN impossible. We present CGBayesNets, a BN package focused around prediction of a clinical phenotype from mixed discrete and continuous variables, which fills these gaps. CGBayesNets implements Bayesian likelihood and inference algorithms for the conditional Gaussian Bayesian network (CGBNs) formalism, one appropriate for predicting an outcome of interest from, e.g., multimodal genomic data. We provide four different network learning algorithms, each making a different tradeoff between computational cost and network likelihood. CGBayesNets provides a full suite of functions for model exploration and verification, including cross validation, bootstrapping, and AUC manipulation. We highlight several results obtained previously with CGBayesNets, including predictive models of wood properties from tree genomics, leukemia subtype classification from mixed genomic data, and robust prediction of intensive care unit mortality outcomes from metabolomic profiles. We also provide detailed example analysis on public metabolomic and gene expression datasets. CGBayesNets is implemented in MATLAB and available as MATLAB source code, under an Open Source license and anonymous download at http://www.cgbayesnets.com. PMID:24922310

CGBayesNets: conditional Gaussian Bayesian network learning and inference with mixed discrete and continuous data.

PubMed

McGeachie, Michael J; Chang, Hsun-Hsien; Weiss, Scott T

2014-06-01

Bayesian Networks (BN) have been a popular predictive modeling formalism in bioinformatics, but their application in modern genomics has been slowed by an inability to cleanly handle domains with mixed discrete and continuous variables. Existing free BN software packages either discretize continuous variables, which can lead to information loss, or do not include inference routines, which makes prediction with the BN impossible. We present CGBayesNets, a BN package focused around prediction of a clinical phenotype from mixed discrete and continuous variables, which fills these gaps. CGBayesNets implements Bayesian likelihood and inference algorithms for the conditional Gaussian Bayesian network (CGBNs) formalism, one appropriate for predicting an outcome of interest from, e.g., multimodal genomic data. We provide four different network learning algorithms, each making a different tradeoff between computational cost and network likelihood. CGBayesNets provides a full suite of functions for model exploration and verification, including cross validation, bootstrapping, and AUC manipulation. We highlight several results obtained previously with CGBayesNets, including predictive models of wood properties from tree genomics, leukemia subtype classification from mixed genomic data, and robust prediction of intensive care unit mortality outcomes from metabolomic profiles. We also provide detailed example analysis on public metabolomic and gene expression datasets. CGBayesNets is implemented in MATLAB and available as MATLAB source code, under an Open Source license and anonymous download at http://www.cgbayesnets.com.

Protein construct storage: Bayesian variable selection and prediction with mixtures.

PubMed

Clyde, M A; Parmigiani, G

1998-07-01

Determining optimal conditions for protein storage while maintaining a high level of protein activity is an important question in pharmaceutical research. A designed experiment based on a space-filling design was conducted to understand the effects of factors affecting protein storage and to establish optimal storage conditions. Different model-selection strategies to identify important factors may lead to very different answers about optimal conditions. Uncertainty about which factors are important, or model uncertainty, can be a critical issue in decision-making. We use Bayesian variable selection methods for linear models to identify important variables in the protein storage data, while accounting for model uncertainty. We also use the Bayesian framework to build predictions based on a large family of models, rather than an individual model, and to evaluate the probability that certain candidate storage conditions are optimal.

Bayesian random-effect model for predicting outcome fraught with heterogeneity--an illustration with episodes of 44 patients with intractable epilepsy.

PubMed

Yen, A M-F; Liou, H-H; Lin, H-L; Chen, T H-H

2006-01-01

The study aimed to develop a predictive model to deal with data fraught with heterogeneity that cannot be explained by sampling variation or measured covariates. The random-effect Poisson regression model was first proposed to deal with over-dispersion for data fraught with heterogeneity after making allowance for measured covariates. Bayesian acyclic graphic model in conjunction with Markov Chain Monte Carlo (MCMC) technique was then applied to estimate the parameters of both relevant covariates and random effect. Predictive distribution was then generated to compare the predicted with the observed for the Bayesian model with and without random effect. Data from repeated measurement of episodes among 44 patients with intractable epilepsy were used as an illustration. The application of Poisson regression without taking heterogeneity into account to epilepsy data yielded a large value of heterogeneity (heterogeneity factor = 17.90, deviance = 1485, degree of freedom (df) = 83). After taking the random effect into account, the value of heterogeneity factor was greatly reduced (heterogeneity factor = 0.52, deviance = 42.5, df = 81). The Pearson chi2 for the comparison between the expected seizure frequencies and the observed ones at two and three months of the model with and without random effect were 34.27 (p = 1.00) and 1799.90 (p < 0.0001), respectively. The Bayesian acyclic model using the MCMC method was demonstrated to have great potential for disease prediction while data show over-dispersion attributed either to correlated property or to subject-to-subject variability.

Testing Bayesian and heuristic predictions of mass judgments of colliding objects

PubMed Central

Sanborn, Adam N.

2014-01-01

Mass judgments of colliding objects have been used to explore people's understanding of the physical world because they are ecologically relevant, yet people display biases that are most easily explained by a small set of heuristics. Recent work has challenged the heuristic explanation, by producing the same biases from a model that copes with perceptual uncertainty by using Bayesian inference with a prior based on the correct combination rules from Newtonian mechanics (noisy Newton). Here I test the predictions of the leading heuristic model (Gilden and Proffitt, 1989) against the noisy Newton model using a novel manipulation of the standard mass judgment task: making one of the objects invisible post-collision. The noisy Newton model uses the remaining information to predict above-chance performance, while the leading heuristic model predicts chance performance when one or the other final velocity is occluded. An experiment using two different types of occlusion showed better-than-chance performance and response patterns that followed the predictions of the noisy Newton model. The results demonstrate that people can make sensible physical judgments even when information critical for the judgment is missing, and that a Bayesian model can serve as a guide in these situations. Possible algorithmic-level accounts of this task that more closely correspond to the noisy Newton model are explored. PMID:25206345

Comparing spatially varying coefficient models: a case study examining violent crime rates and their relationships to alcohol outlets and illegal drug arrests

NASA Astrophysics Data System (ADS)

Wheeler, David C.; Waller, Lance A.

2009-03-01

In this paper, we compare and contrast a Bayesian spatially varying coefficient process (SVCP) model with a geographically weighted regression (GWR) model for the estimation of the potentially spatially varying regression effects of alcohol outlets and illegal drug activity on violent crime in Houston, Texas. In addition, we focus on the inherent coefficient shrinkage properties of the Bayesian SVCP model as a way to address increased coefficient variance that follows from collinearity in GWR models. We outline the advantages of the Bayesian model in terms of reducing inflated coefficient variance, enhanced model flexibility, and more formal measuring of model uncertainty for prediction. We find spatially varying effects for alcohol outlets and drug violations, but the amount of variation depends on the type of model used. For the Bayesian model, this variation is controllable through the amount of prior influence placed on the variance of the coefficients. For example, the spatial pattern of coefficients is similar for the GWR and Bayesian models when a relatively large prior variance is used in the Bayesian model.

The Bayesian reader: explaining word recognition as an optimal Bayesian decision process.

PubMed

Norris, Dennis

2006-04-01

This article presents a theory of visual word recognition that assumes that, in the tasks of word identification, lexical decision, and semantic categorization, human readers behave as optimal Bayesian decision makers. This leads to the development of a computational model of word recognition, the Bayesian reader. The Bayesian reader successfully simulates some of the most significant data on human reading. The model accounts for the nature of the function relating word frequency to reaction time and identification threshold, the effects of neighborhood density and its interaction with frequency, and the variation in the pattern of neighborhood density effects seen in different experimental tasks. Both the general behavior of the model and the way the model predicts different patterns of results in different tasks follow entirely from the assumption that human readers approximate optimal Bayesian decision makers. ((c) 2006 APA, all rights reserved).

A cost minimisation and Bayesian inference model predicts startle reflex modulation across species.

PubMed

Bach, Dominik R

2015-04-07

In many species, rapid defensive reflexes are paramount to escaping acute danger. These reflexes are modulated by the state of the environment. This is exemplified in fear-potentiated startle, a more vigorous startle response during conditioned anticipation of an unrelated threatening event. Extant explanations of this phenomenon build on descriptive models of underlying psychological states, or neural processes. Yet, they fail to predict invigorated startle during reward anticipation and instructed attention, and do not explain why startle reflex modulation evolved. Here, we fill this lacuna by developing a normative cost minimisation model based on Bayesian optimality principles. This model predicts the observed pattern of startle modification by rewards, punishments, instructed attention, and several other states. Moreover, the mathematical formalism furnishes predictions that can be tested experimentally. Comparing the model with existing data suggests a specific neural implementation of the underlying computations which yields close approximations to the optimal solution under most circumstances. This analysis puts startle modification into the framework of Bayesian decision theory and predictive coding, and illustrates the importance of an adaptive perspective to interpret defensive behaviour across species. Copyright © 2015 The Author. Published by Elsevier Ltd.. All rights reserved.

Construction of monitoring model and algorithm design on passenger security during shipping based on improved Bayesian network.

PubMed

Wang, Jiali; Zhang, Qingnian; Ji, Wenfeng

2014-01-01

A large number of data is needed by the computation of the objective Bayesian network, but the data is hard to get in actual computation. The calculation method of Bayesian network was improved in this paper, and the fuzzy-precise Bayesian network was obtained. Then, the fuzzy-precise Bayesian network was used to reason Bayesian network model when the data is limited. The security of passengers during shipping is affected by various factors, and it is hard to predict and control. The index system that has the impact on the passenger safety during shipping was established on basis of the multifield coupling theory in this paper. Meanwhile, the fuzzy-precise Bayesian network was applied to monitor the security of passengers in the shipping process. The model was applied to monitor the passenger safety during shipping of a shipping company in Hainan, and the effectiveness of this model was examined. This research work provides guidance for guaranteeing security of passengers during shipping.

Construction of Monitoring Model and Algorithm Design on Passenger Security during Shipping Based on Improved Bayesian Network

PubMed Central

Wang, Jiali; Zhang, Qingnian; Ji, Wenfeng

2014-01-01

A large number of data is needed by the computation of the objective Bayesian network, but the data is hard to get in actual computation. The calculation method of Bayesian network was improved in this paper, and the fuzzy-precise Bayesian network was obtained. Then, the fuzzy-precise Bayesian network was used to reason Bayesian network model when the data is limited. The security of passengers during shipping is affected by various factors, and it is hard to predict and control. The index system that has the impact on the passenger safety during shipping was established on basis of the multifield coupling theory in this paper. Meanwhile, the fuzzy-precise Bayesian network was applied to monitor the security of passengers in the shipping process. The model was applied to monitor the passenger safety during shipping of a shipping company in Hainan, and the effectiveness of this model was examined. This research work provides guidance for guaranteeing security of passengers during shipping. PMID:25254227

Predicting Software Suitability Using a Bayesian Belief Network

NASA Technical Reports Server (NTRS)

Beaver, Justin M.; Schiavone, Guy A.; Berrios, Joseph S.

2005-01-01

The ability to reliably predict the end quality of software under development presents a significant advantage for a development team. It provides an opportunity to address high risk components earlier in the development life cycle, when their impact is minimized. This research proposes a model that captures the evolution of the quality of a software product, and provides reliable forecasts of the end quality of the software being developed in terms of product suitability. Development team skill, software process maturity, and software problem complexity are hypothesized as driving factors of software product quality. The cause-effect relationships between these factors and the elements of software suitability are modeled using Bayesian Belief Networks, a machine learning method. This research presents a Bayesian Network for software quality, and the techniques used to quantify the factors that influence and represent software quality. The developed model is found to be effective in predicting the end product quality of small-scale software development efforts.

Performance of two predictive uncertainty estimation approaches for conceptual Rainfall-Runoff Model: Bayesian Joint Inference and Hydrologic Uncertainty Post-processing

NASA Astrophysics Data System (ADS)

Hernández-López, Mario R.; Romero-Cuéllar, Jonathan; Camilo Múnera-Estrada, Juan; Coccia, Gabriele; Francés, Félix

2017-04-01

It is noticeably important to emphasize the role of uncertainty particularly when the model forecasts are used to support decision-making and water management. This research compares two approaches for the evaluation of the predictive uncertainty in hydrological modeling. First approach is the Bayesian Joint Inference of hydrological and error models. Second approach is carried out through the Model Conditional Processor using the Truncated Normal Distribution in the transformed space. This comparison is focused on the predictive distribution reliability. The case study is applied to two basins included in the Model Parameter Estimation Experiment (MOPEX). These two basins, which have different hydrological complexity, are the French Broad River (North Carolina) and the Guadalupe River (Texas). The results indicate that generally, both approaches are able to provide similar predictive performances. However, the differences between them can arise in basins with complex hydrology (e.g. ephemeral basins). This is because obtained results with Bayesian Joint Inference are strongly dependent on the suitability of the hypothesized error model. Similarly, the results in the case of the Model Conditional Processor are mainly influenced by the selected model of tails or even by the selected full probability distribution model of the data in the real space, and by the definition of the Truncated Normal Distribution in the transformed space. In summary, the different hypotheses that the modeler choose on each of the two approaches are the main cause of the different results. This research also explores a proper combination of both methodologies which could be useful to achieve less biased hydrological parameter estimation. For this approach, firstly the predictive distribution is obtained through the Model Conditional Processor. Secondly, this predictive distribution is used to derive the corresponding additive error model which is employed for the hydrological parameter estimation with the Bayesian Joint Inference methodology.

A Bayesian Framework for False Belief Reasoning in Children: A Rational Integration of Theory-Theory and Simulation Theory

PubMed Central

Asakura, Nobuhiko; Inui, Toshio

2016-01-01

Two apparently contrasting theories have been proposed to account for the development of children's theory of mind (ToM): theory-theory and simulation theory. We present a Bayesian framework that rationally integrates both theories for false belief reasoning. This framework exploits two internal models for predicting the belief states of others: one of self and one of others. These internal models are responsible for simulation-based and theory-based reasoning, respectively. The framework further takes into account empirical studies of a developmental ToM scale (e.g., Wellman and Liu, 2004): developmental progressions of various mental state understandings leading up to false belief understanding. By representing the internal models and their interactions as a causal Bayesian network, we formalize the model of children's false belief reasoning as probabilistic computations on the Bayesian network. This model probabilistically weighs and combines the two internal models and predicts children's false belief ability as a multiplicative effect of their early-developed abilities to understand the mental concepts of diverse beliefs and knowledge access. Specifically, the model predicts that children's proportion of correct responses on a false belief task can be closely approximated as the product of their proportions correct on the diverse belief and knowledge access tasks. To validate this prediction, we illustrate that our model provides good fits to a variety of ToM scale data for preschool children. We discuss the implications and extensions of our model for a deeper understanding of developmental progressions of children's ToM abilities. PMID:28082941

A Bayesian Framework for False Belief Reasoning in Children: A Rational Integration of Theory-Theory and Simulation Theory.

PubMed

Asakura, Nobuhiko; Inui, Toshio

2016-01-01

Two apparently contrasting theories have been proposed to account for the development of children's theory of mind (ToM): theory-theory and simulation theory. We present a Bayesian framework that rationally integrates both theories for false belief reasoning. This framework exploits two internal models for predicting the belief states of others: one of self and one of others. These internal models are responsible for simulation-based and theory-based reasoning, respectively. The framework further takes into account empirical studies of a developmental ToM scale (e.g., Wellman and Liu, 2004): developmental progressions of various mental state understandings leading up to false belief understanding. By representing the internal models and their interactions as a causal Bayesian network, we formalize the model of children's false belief reasoning as probabilistic computations on the Bayesian network. This model probabilistically weighs and combines the two internal models and predicts children's false belief ability as a multiplicative effect of their early-developed abilities to understand the mental concepts of diverse beliefs and knowledge access. Specifically, the model predicts that children's proportion of correct responses on a false belief task can be closely approximated as the product of their proportions correct on the diverse belief and knowledge access tasks. To validate this prediction, we illustrate that our model provides good fits to a variety of ToM scale data for preschool children. We discuss the implications and extensions of our model for a deeper understanding of developmental progressions of children's ToM abilities.

Applying Bayesian Modeling and Receiver Operating Characteristic Methodologies for Test Utility Analysis

ERIC Educational Resources Information Center

Wang, Qiu; Diemer, Matthew A.; Maier, Kimberly S.

2013-01-01

This study integrated Bayesian hierarchical modeling and receiver operating characteristic analysis (BROCA) to evaluate how interest strength (IS) and interest differentiation (ID) predicted low–socioeconomic status (SES) youth's interest-major congruence (IMC). Using large-scale Kuder Career Search online-assessment data, this study fit three…

Predicting coastal cliff erosion using a Bayesian probabilistic model

USGS Publications Warehouse

Hapke, Cheryl J.; Plant, Nathaniel G.

2010-01-01

Regional coastal cliff retreat is difficult to model due to the episodic nature of failures and the along-shore variability of retreat events. There is a growing demand, however, for predictive models that can be used to forecast areas vulnerable to coastal erosion hazards. Increasingly, probabilistic models are being employed that require data sets of high temporal density to define the joint probability density function that relates forcing variables (e.g. wave conditions) and initial conditions (e.g. cliff geometry) to erosion events. In this study we use a multi-parameter Bayesian network to investigate correlations between key variables that control and influence variations in cliff retreat processes. The network uses Bayesian statistical methods to estimate event probabilities using existing observations. Within this framework, we forecast the spatial distribution of cliff retreat along two stretches of cliffed coast in Southern California. The input parameters are the height and slope of the cliff, a descriptor of material strength based on the dominant cliff-forming lithology, and the long-term cliff erosion rate that represents prior behavior. The model is forced using predicted wave impact hours. Results demonstrate that the Bayesian approach is well-suited to the forward modeling of coastal cliff retreat, with the correct outcomes forecast in 70–90% of the modeled transects. The model also performs well in identifying specific locations of high cliff erosion, thus providing a foundation for hazard mapping. This approach can be employed to predict cliff erosion at time-scales ranging from storm events to the impacts of sea-level rise at the century-scale.

Improved prediction of tacrolimus concentrations early after kidney transplantation using theory-based pharmacokinetic modelling.

PubMed

Størset, Elisabet; Holford, Nick; Hennig, Stefanie; Bergmann, Troels K; Bergan, Stein; Bremer, Sara; Åsberg, Anders; Midtvedt, Karsten; Staatz, Christine E

2014-09-01

The aim was to develop a theory-based population pharmacokinetic model of tacrolimus in adult kidney transplant recipients and to externally evaluate this model and two previous empirical models. Data were obtained from 242 patients with 3100 tacrolimus whole blood concentrations. External evaluation was performed by examining model predictive performance using Bayesian forecasting. Pharmacokinetic disposition parameters were estimated based on tacrolimus plasma concentrations, predicted from whole blood concentrations, haematocrit and literature values for tacrolimus binding to red blood cells. Disposition parameters were allometrically scaled to fat free mass. Tacrolimus whole blood clearance/bioavailability standardized to haematocrit of 45% and fat free mass of 60 kg was estimated to be 16.1 l h−1 [95% CI 12.6, 18.0 l h−1]. Tacrolimus clearance was 30% higher (95% CI 13, 46%) and bioavailability 18% lower (95% CI 2, 29%) in CYP3A5 expressers compared with non-expressers. An Emax model described decreasing tacrolimus bioavailability with increasing prednisolone dose. The theory-based model was superior to the empirical models during external evaluation displaying a median prediction error of −1.2% (95% CI −3.0, 0.1%). Based on simulation, Bayesian forecasting led to 65% (95% CI 62, 68%) of patients achieving a tacrolimus average steady-state concentration within a suggested acceptable range. A theory-based population pharmacokinetic model was superior to two empirical models for prediction of tacrolimus concentrations and seemed suitable for Bayesian prediction of tacrolimus doses early after kidney transplantation.

Bayesian cross-validation for model evaluation and selection, with application to the North American Breeding Bird Survey

USGS Publications Warehouse

Link, William; Sauer, John R.

2016-01-01

The analysis of ecological data has changed in two important ways over the last 15 years. The development and easy availability of Bayesian computational methods has allowed and encouraged the fitting of complex hierarchical models. At the same time, there has been increasing emphasis on acknowledging and accounting for model uncertainty. Unfortunately, the ability to fit complex models has outstripped the development of tools for model selection and model evaluation: familiar model selection tools such as Akaike's information criterion and the deviance information criterion are widely known to be inadequate for hierarchical models. In addition, little attention has been paid to the evaluation of model adequacy in context of hierarchical modeling, i.e., to the evaluation of fit for a single model. In this paper, we describe Bayesian cross-validation, which provides tools for model selection and evaluation. We describe the Bayesian predictive information criterion and a Bayesian approximation to the BPIC known as the Watanabe-Akaike information criterion. We illustrate the use of these tools for model selection, and the use of Bayesian cross-validation as a tool for model evaluation, using three large data sets from the North American Breeding Bird Survey.

Predicting uncertainty in future marine ice sheet volume using Bayesian statistical methods

NASA Astrophysics Data System (ADS)

Davis, A. D.

2015-12-01

The marine ice instability can trigger rapid retreat of marine ice streams. Recent observations suggest that marine ice systems in West Antarctica have begun retreating. However, unknown ice dynamics, computationally intensive mathematical models, and uncertain parameters in these models make predicting retreat rate and ice volume difficult. In this work, we fuse current observational data with ice stream/shelf models to develop probabilistic predictions of future grounded ice sheet volume. Given observational data (e.g., thickness, surface elevation, and velocity) and a forward model that relates uncertain parameters (e.g., basal friction and basal topography) to these observations, we use a Bayesian framework to define a posterior distribution over the parameters. A stochastic predictive model then propagates uncertainties in these parameters to uncertainty in a particular quantity of interest (QoI)---here, the volume of grounded ice at a specified future time. While the Bayesian approach can in principle characterize the posterior predictive distribution of the QoI, the computational cost of both the forward and predictive models makes this effort prohibitively expensive. To tackle this challenge, we introduce a new Markov chain Monte Carlo method that constructs convergent approximations of the QoI target density in an online fashion, yielding accurate characterizations of future ice sheet volume at significantly reduced computational cost.Our second goal is to attribute uncertainty in these Bayesian predictions to uncertainties in particular parameters. Doing so can help target data collection, for the purpose of constraining the parameters that contribute most strongly to uncertainty in the future volume of grounded ice. For instance, smaller uncertainties in parameters to which the QoI is highly sensitive may account for more variability in the prediction than larger uncertainties in parameters to which the QoI is less sensitive. We use global sensitivity analysis to help answer this question, and make the computation of sensitivity indices computationally tractable using a combination of polynomial chaos and Monte Carlo techniques.

Hip fracture in the elderly: a re-analysis of the EPIDOS study with causal Bayesian networks.

PubMed

Caillet, Pascal; Klemm, Sarah; Ducher, Michel; Aussem, Alexandre; Schott, Anne-Marie

2015-01-01

Hip fractures commonly result in permanent disability, institutionalization or death in elderly. Existing hip-fracture predicting tools are underused in clinical practice, partly due to their lack of intuitive interpretation. By use of a graphical layer, Bayesian network models could increase the attractiveness of fracture prediction tools. Our aim was to study the potential contribution of a causal Bayesian network in this clinical setting. A logistic regression was performed as a standard control approach to check the robustness of the causal Bayesian network approach. EPIDOS is a multicenter study, conducted in an ambulatory care setting in five French cities between 1992 and 1996 and updated in 2010. The study included 7598 women aged 75 years or older, in which fractures were assessed quarterly during 4 years. A causal Bayesian network and a logistic regression were performed on EPIDOS data to describe major variables involved in hip fractures occurrences. Both models had similar association estimations and predictive performances. They detected gait speed and mineral bone density as variables the most involved in the fracture process. The causal Bayesian network showed that gait speed and bone mineral density were directly connected to fracture and seem to mediate the influence of all the other variables included in our model. The logistic regression approach detected multiple interactions involving psychotropic drug use, age and bone mineral density. Both approaches retrieved similar variables as predictors of hip fractures. However, Bayesian network highlighted the whole web of relation between the variables involved in the analysis, suggesting a possible mechanism leading to hip fracture. According to the latter results, intervention focusing concomitantly on gait speed and bone mineral density may be necessary for an optimal prevention of hip fracture occurrence in elderly people.

Assimilating multi-source uncertainties of a parsimonious conceptual hydrological model using hierarchical Bayesian modeling

Treesearch

Wei Wu; James Clark; James Vose

2010-01-01

Hierarchical Bayesian (HB) modeling allows for multiple sources of uncertainty by factoring complex relationships into conditional distributions that can be used to draw inference and make predictions. We applied an HB model to estimate the parameters and state variables of a parsimonious hydrological model – GR4J – by coherently assimilating the uncertainties from the...

Nuclear charge radii: density functional theory meets Bayesian neural networks

NASA Astrophysics Data System (ADS)

Utama, R.; Chen, Wei-Chia; Piekarewicz, J.

2016-11-01

The distribution of electric charge in atomic nuclei is fundamental to our understanding of the complex nuclear dynamics and a quintessential observable to validate nuclear structure models. The aim of this study is to explore a novel approach that combines sophisticated models of nuclear structure with Bayesian neural networks (BNN) to generate predictions for the charge radii of thousands of nuclei throughout the nuclear chart. A class of relativistic energy density functionals is used to provide robust predictions for nuclear charge radii. In turn, these predictions are refined through Bayesian learning for a neural network that is trained using residuals between theoretical predictions and the experimental data. Although predictions obtained with density functional theory provide a fairly good description of experiment, our results show significant improvement (better than 40%) after BNN refinement. Moreover, these improved results for nuclear charge radii are supplemented with theoretical error bars. We have successfully demonstrated the ability of the BNN approach to significantly increase the accuracy of nuclear models in the predictions of nuclear charge radii. However, as many before us, we failed to uncover the underlying physics behind the intriguing behavior of charge radii along the calcium isotopic chain.

A Bayesian network model for predicting pregnancy after in vitro fertilization.

PubMed

Corani, G; Magli, C; Giusti, A; Gianaroli, L; Gambardella, L M

2013-11-01

We present a Bayesian network model for predicting the outcome of in vitro fertilization (IVF). The problem is characterized by a particular missingness process; we propose a simple but effective averaging approach which improves parameter estimates compared to the traditional MAP estimation. We present results with generated data and the analysis of a real data set. Moreover, we assess by means of a simulation study the effectiveness of the model in supporting the selection of the embryos to be transferred. © 2013 Elsevier Ltd. All rights reserved.

Perceptual decision making: drift-diffusion model is equivalent to a Bayesian model

PubMed Central

Bitzer, Sebastian; Park, Hame; Blankenburg, Felix; Kiebel, Stefan J.

2014-01-01

Behavioral data obtained with perceptual decision making experiments are typically analyzed with the drift-diffusion model. This parsimonious model accumulates noisy pieces of evidence toward a decision bound to explain the accuracy and reaction times of subjects. Recently, Bayesian models have been proposed to explain how the brain extracts information from noisy input as typically presented in perceptual decision making tasks. It has long been known that the drift-diffusion model is tightly linked with such functional Bayesian models but the precise relationship of the two mechanisms was never made explicit. Using a Bayesian model, we derived the equations which relate parameter values between these models. In practice we show that this equivalence is useful when fitting multi-subject data. We further show that the Bayesian model suggests different decision variables which all predict equal responses and discuss how these may be discriminated based on neural correlates of accumulated evidence. In addition, we discuss extensions to the Bayesian model which would be difficult to derive for the drift-diffusion model. We suggest that these and other extensions may be highly useful for deriving new experiments which test novel hypotheses. PMID:24616689

Predicting Rotator Cuff Tears Using Data Mining and Bayesian Likelihood Ratios

PubMed Central

Lu, Hsueh-Yi; Huang, Chen-Yuan; Su, Chwen-Tzeng; Lin, Chen-Chiang

2014-01-01

Objectives Rotator cuff tear is a common cause of shoulder diseases. Correct diagnosis of rotator cuff tears can save patients from further invasive, costly and painful tests. This study used predictive data mining and Bayesian theory to improve the accuracy of diagnosing rotator cuff tears by clinical examination alone. Methods In this retrospective study, 169 patients who had a preliminary diagnosis of rotator cuff tear on the basis of clinical evaluation followed by confirmatory MRI between 2007 and 2011 were identified. MRI was used as a reference standard to classify rotator cuff tears. The predictor variable was the clinical assessment results, which consisted of 16 attributes. This study employed 2 data mining methods (ANN and the decision tree) and a statistical method (logistic regression) to classify the rotator cuff diagnosis into “tear” and “no tear” groups. Likelihood ratio and Bayesian theory were applied to estimate the probability of rotator cuff tears based on the results of the prediction models. Results Our proposed data mining procedures outperformed the classic statistical method. The correction rate, sensitivity, specificity and area under the ROC curve of predicting a rotator cuff tear were statistical better in the ANN and decision tree models compared to logistic regression. Based on likelihood ratios derived from our prediction models, Fagan's nomogram could be constructed to assess the probability of a patient who has a rotator cuff tear using a pretest probability and a prediction result (tear or no tear). Conclusions Our predictive data mining models, combined with likelihood ratios and Bayesian theory, appear to be good tools to classify rotator cuff tears as well as determine the probability of the presence of the disease to enhance diagnostic decision making for rotator cuff tears. PMID:24733553

On the use of posterior predictive probabilities and prediction uncertainty to tailor informative sampling for parasitological surveillance in livestock.

PubMed

Musella, Vincenzo; Rinaldi, Laura; Lagazio, Corrado; Cringoli, Giuseppe; Biggeri, Annibale; Catelan, Dolores

2014-09-15

Model-based geostatistics and Bayesian approaches are appropriate in the context of Veterinary Epidemiology when point data have been collected by valid study designs. The aim is to predict a continuous infection risk surface. Little work has been done on the use of predictive infection probabilities at farm unit level. In this paper we show how to use predictive infection probability and related uncertainty from a Bayesian kriging model to draw a informative samples from the 8794 geo-referenced sheep farms of the Campania region (southern Italy). Parasitological data come from a first cross-sectional survey carried out to study the spatial distribution of selected helminths in sheep farms. A grid sampling was performed to select the farms for coprological examinations. Faecal samples were collected for 121 sheep farms and the presence of 21 different helminths were investigated using the FLOTAC technique. The 21 responses are very different in terms of geographical distribution and prevalence of infection. The observed prevalence range is from 0.83% to 96.69%. The distributions of the posterior predictive probabilities for all the 21 parasites are very heterogeneous. We show how the results of the Bayesian kriging model can be used to plan a second wave survey. Several alternatives can be chosen depending on the purposes of the second survey: weight by posterior predictive probabilities, their uncertainty or combining both information. The proposed Bayesian kriging model is simple, and the proposed samping strategy represents a useful tool to address targeted infection control treatments and surbveillance campaigns. It is easily extendable to other fields of research. Copyright © 2014 Elsevier B.V. All rights reserved.

DOE Office of Scientific and Technical Information (OSTI.GOV)

Vrugt, Jasper A; Robinson, Bruce A; Ter Braak, Cajo J F

In recent years, a strong debate has emerged in the hydrologic literature regarding what constitutes an appropriate framework for uncertainty estimation. Particularly, there is strong disagreement whether an uncertainty framework should have its roots within a proper statistical (Bayesian) context, or whether such a framework should be based on a different philosophy and implement informal measures and weaker inference to summarize parameter and predictive distributions. In this paper, we compare a formal Bayesian approach using Markov Chain Monte Carlo (MCMC) with generalized likelihood uncertainty estimation (GLUE) for assessing uncertainty in conceptual watershed modeling. Our formal Bayesian approach is implemented usingmore » the recently developed differential evolution adaptive metropolis (DREAM) MCMC scheme with a likelihood function that explicitly considers model structural, input and parameter uncertainty. Our results demonstrate that DREAM and GLUE can generate very similar estimates of total streamflow uncertainty. This suggests that formal and informal Bayesian approaches have more common ground than the hydrologic literature and ongoing debate might suggest. The main advantage of formal approaches is, however, that they attempt to disentangle the effect of forcing, parameter and model structural error on total predictive uncertainty. This is key to improving hydrologic theory and to better understand and predict the flow of water through catchments.« less

Trust from the past: Bayesian Personalized Ranking based Link Prediction in Knowledge Graphs

DOE Office of Scientific and Technical Information (OSTI.GOV)

Zhang, Baichuan; Choudhury, Sutanay; Al-Hasan, Mohammad

2016-02-01

Estimating the confidence for a link is a critical task for Knowledge Graph construction. Link prediction, or predicting the likelihood of a link in a knowledge graph based on prior state is a key research direction within this area. We propose a Latent Feature Embedding based link recommendation model for prediction task and utilize Bayesian Personalized Ranking based optimization technique for learning models for each predicate. Experimental results on large-scale knowledge bases such as YAGO2 show that our approach achieves substantially higher performance than several state-of-art approaches. Furthermore, we also study the performance of the link prediction algorithm in termsmore » of topological properties of the Knowledge Graph and present a linear regression model to reason about its expected level of accuracy.« less

A Bayesian approach for parameter estimation and prediction using a computationally intensive model

DOE PAGES

Higdon, Dave; McDonnell, Jordan D.; Schunck, Nicolas; ...

2015-02-05

Bayesian methods have been successful in quantifying uncertainty in physics-based problems in parameter estimation and prediction. In these cases, physical measurements y are modeled as the best fit of a physics-based modelmore » $$\\eta (\\theta )$$, where θ denotes the uncertain, best input setting. Hence the statistical model is of the form $$y=\\eta (\\theta )+\\epsilon ,$$ where $$\\epsilon $$ accounts for measurement, and possibly other, error sources. When nonlinearity is present in $$\\eta (\\cdot )$$, the resulting posterior distribution for the unknown parameters in the Bayesian formulation is typically complex and nonstandard, requiring computationally demanding computational approaches such as Markov chain Monte Carlo (MCMC) to produce multivariate draws from the posterior. Although generally applicable, MCMC requires thousands (or even millions) of evaluations of the physics model $$\\eta (\\cdot )$$. This requirement is problematic if the model takes hours or days to evaluate. To overcome this computational bottleneck, we present an approach adapted from Bayesian model calibration. This approach combines output from an ensemble of computational model runs with physical measurements, within a statistical formulation, to carry out inference. A key component of this approach is a statistical response surface, or emulator, estimated from the ensemble of model runs. We demonstrate this approach with a case study in estimating parameters for a density functional theory model, using experimental mass/binding energy measurements from a collection of atomic nuclei. Lastly, we also demonstrate how this approach produces uncertainties in predictions for recent mass measurements obtained at Argonne National Laboratory.« less

Model selection and assessment for multi­-species occupancy models

USGS Publications Warehouse

Broms, Kristin M.; Hooten, Mevin B.; Fitzpatrick, Ryan M.

2016-01-01

While multi-species occupancy models (MSOMs) are emerging as a popular method for analyzing biodiversity data, formal checking and validation approaches for this class of models have lagged behind. Concurrent with the rise in application of MSOMs among ecologists, a quiet regime shift is occurring in Bayesian statistics where predictive model comparison approaches are experiencing a resurgence. Unlike single-species occupancy models that use integrated likelihoods, MSOMs are usually couched in a Bayesian framework and contain multiple levels. Standard model checking and selection methods are often unreliable in this setting and there is only limited guidance in the ecological literature for this class of models. We examined several different contemporary Bayesian hierarchical approaches for checking and validating MSOMs and applied these methods to a freshwater aquatic study system in Colorado, USA, to better understand the diversity and distributions of plains fishes. Our findings indicated distinct differences among model selection approaches, with cross-validation techniques performing the best in terms of prediction.

Bayesian Integration of Information in Hippocampal Place Cells

PubMed Central

Madl, Tamas; Franklin, Stan; Chen, Ke; Montaldi, Daniela; Trappl, Robert

2014-01-01

Accurate spatial localization requires a mechanism that corrects for errors, which might arise from inaccurate sensory information or neuronal noise. In this paper, we propose that Hippocampal place cells might implement such an error correction mechanism by integrating different sources of information in an approximately Bayes-optimal fashion. We compare the predictions of our model with physiological data from rats. Our results suggest that useful predictions regarding the firing fields of place cells can be made based on a single underlying principle, Bayesian cue integration, and that such predictions are possible using a remarkably small number of model parameters. PMID:24603429

Bayesian analysis of rare events

NASA Astrophysics Data System (ADS)

Straub, Daniel; Papaioannou, Iason; Betz, Wolfgang

2016-06-01

In many areas of engineering and science there is an interest in predicting the probability of rare events, in particular in applications related to safety and security. Increasingly, such predictions are made through computer models of physical systems in an uncertainty quantification framework. Additionally, with advances in IT, monitoring and sensor technology, an increasing amount of data on the performance of the systems is collected. This data can be used to reduce uncertainty, improve the probability estimates and consequently enhance the management of rare events and associated risks. Bayesian analysis is the ideal method to include the data into the probabilistic model. It ensures a consistent probabilistic treatment of uncertainty, which is central in the prediction of rare events, where extrapolation from the domain of observation is common. We present a framework for performing Bayesian updating of rare event probabilities, termed BUS. It is based on a reinterpretation of the classical rejection-sampling approach to Bayesian analysis, which enables the use of established methods for estimating probabilities of rare events. By drawing upon these methods, the framework makes use of their computational efficiency. These methods include the First-Order Reliability Method (FORM), tailored importance sampling (IS) methods and Subset Simulation (SuS). In this contribution, we briefly review these methods in the context of the BUS framework and investigate their applicability to Bayesian analysis of rare events in different settings. We find that, for some applications, FORM can be highly efficient and is surprisingly accurate, enabling Bayesian analysis of rare events with just a few model evaluations. In a general setting, BUS implemented through IS and SuS is more robust and flexible.

Probabilistic prediction of barrier-island response to hurricanes

USGS Publications Warehouse

Plant, Nathaniel G.; Stockdon, Hilary F.

2012-01-01

Prediction of barrier-island response to hurricane attack is important for assessing the vulnerability of communities, infrastructure, habitat, and recreational assets to the impacts of storm surge, waves, and erosion. We have demonstrated that a conceptual model intended to make qualitative predictions of the type of beach response to storms (e.g., beach erosion, dune erosion, dune overwash, inundation) can be reformulated in a Bayesian network to make quantitative predictions of the morphologic response. In an application of this approach at Santa Rosa Island, FL, predicted dune-crest elevation changes in response to Hurricane Ivan explained about 20% to 30% of the observed variance. An extended Bayesian network based on the original conceptual model, which included dune elevations, storm surge, and swash, but with the addition of beach and dune widths as input variables, showed improved skill compared to the original model, explaining 70% of dune elevation change variance and about 60% of dune and shoreline position change variance. This probabilistic approach accurately represented prediction uncertainty (measured with the log likelihood ratio), and it outperformed the baseline prediction (i.e., the prior distribution based on the observations). Finally, sensitivity studies demonstrated that degrading the resolution of the Bayesian network or removing data from the calibration process reduced the skill of the predictions by 30% to 40%. The reduction in skill did not change conclusions regarding the relative importance of the input variables, and the extended model's skill always outperformed the original model.

A Bayesian hierarchical model with spatial variable selection: the effect of weather on insurance claims

PubMed Central

Scheel, Ida; Ferkingstad, Egil; Frigessi, Arnoldo; Haug, Ola; Hinnerichsen, Mikkel; Meze-Hausken, Elisabeth

2013-01-01

Climate change will affect the insurance industry. We develop a Bayesian hierarchical statistical approach to explain and predict insurance losses due to weather events at a local geographic scale. The number of weather-related insurance claims is modelled by combining generalized linear models with spatially smoothed variable selection. Using Gibbs sampling and reversible jump Markov chain Monte Carlo methods, this model is fitted on daily weather and insurance data from each of the 319 municipalities which constitute southern and central Norway for the period 1997–2006. Precise out-of-sample predictions validate the model. Our results show interesting regional patterns in the effect of different weather covariates. In addition to being useful for insurance pricing, our model can be used for short-term predictions based on weather forecasts and for long-term predictions based on downscaled climate models. PMID:23396890

Predicting Drug Safety and Communicating Risk: Benefits of a Bayesian Approach.

PubMed

Lazic, Stanley E; Edmunds, Nicholas; Pollard, Christopher E

2018-03-01

Drug toxicity is a major source of attrition in drug discovery and development. Pharmaceutical companies routinely use preclinical data to predict clinical outcomes and continue to invest in new assays to improve predictions. However, there are many open questions about how to make the best use of available data, combine diverse data, quantify risk, and communicate risk and uncertainty to enable good decisions. The costs of suboptimal decisions are clear: resources are wasted and patients may be put at risk. We argue that Bayesian methods provide answers to all of these problems and use hERG-mediated QT prolongation as a case study. Benefits of Bayesian machine learning models include intuitive probabilistic statements of risk that incorporate all sources of uncertainty, the option to include diverse data and external information, and visualizations that have a clear link between the output from a statistical model and what this means for risk. Furthermore, Bayesian methods are easy to use with modern software, making their adoption for safety screening straightforward. We include R and Python code to encourage the adoption of these methods.

Bayesian prediction of future ice sheet volume using local approximation Markov chain Monte Carlo methods

NASA Astrophysics Data System (ADS)

Davis, A. D.; Heimbach, P.; Marzouk, Y.

2017-12-01

We develop a Bayesian inverse modeling framework for predicting future ice sheet volume with associated formal uncertainty estimates. Marine ice sheets are drained by fast-flowing ice streams, which we simulate using a flowline model. Flowline models depend on geometric parameters (e.g., basal topography), parameterized physical processes (e.g., calving laws and basal sliding), and climate parameters (e.g., surface mass balance), most of which are unknown or uncertain. Given observations of ice surface velocity and thickness, we define a Bayesian posterior distribution over static parameters, such as basal topography. We also define a parameterized distribution over variable parameters, such as future surface mass balance, which we assume are not informed by the data. Hyperparameters are used to represent climate change scenarios, and sampling their distributions mimics internal variation. For example, a warming climate corresponds to increasing mean surface mass balance but an individual sample may have periods of increasing or decreasing surface mass balance. We characterize the predictive distribution of ice volume by evaluating the flowline model given samples from the posterior distribution and the distribution over variable parameters. Finally, we determine the effect of climate change on future ice sheet volume by investigating how changing the hyperparameters affects the predictive distribution. We use state-of-the-art Bayesian computation to address computational feasibility. Characterizing the posterior distribution (using Markov chain Monte Carlo), sampling the full range of variable parameters and evaluating the predictive model is prohibitively expensive. Furthermore, the required resolution of the inferred basal topography may be very high, which is often challenging for sampling methods. Instead, we leverage regularity in the predictive distribution to build a computationally cheaper surrogate over the low dimensional quantity of interest (future ice sheet volume). Continual surrogate refinement guarantees asymptotic sampling from the predictive distribution. Directly characterizing the predictive distribution in this way allows us to assess the ice sheet's sensitivity to climate variability and change.

Using Bayesian analysis in repeated preclinical in vivo studies for a more effective use of animals.

PubMed

Walley, Rosalind; Sherington, John; Rastrick, Joe; Detrait, Eric; Hanon, Etienne; Watt, Gillian

2016-05-01

Whilst innovative Bayesian approaches are increasingly used in clinical studies, in the preclinical area Bayesian methods appear to be rarely used in the reporting of pharmacology data. This is particularly surprising in the context of regularly repeated in vivo studies where there is a considerable amount of data from historical control groups, which has potential value. This paper describes our experience with introducing Bayesian analysis for such studies using a Bayesian meta-analytic predictive approach. This leads naturally either to an informative prior for a control group as part of a full Bayesian analysis of the next study or using a predictive distribution to replace a control group entirely. We use quality control charts to illustrate study-to-study variation to the scientists and describe informative priors in terms of their approximate effective numbers of animals. We describe two case studies of animal models: the lipopolysaccharide-induced cytokine release model used in inflammation and the novel object recognition model used to screen cognitive enhancers, both of which show the advantage of a Bayesian approach over the standard frequentist analysis. We conclude that using Bayesian methods in stable repeated in vivo studies can result in a more effective use of animals, either by reducing the total number of animals used or by increasing the precision of key treatment differences. This will lead to clearer results and supports the "3Rs initiative" to Refine, Reduce and Replace animals in research. Copyright © 2016 John Wiley & Sons, Ltd. Copyright © 2016 John Wiley & Sons, Ltd.

Risk prediction model for knee pain in the Nottingham community: a Bayesian modelling approach.

PubMed

Fernandes, G S; Bhattacharya, A; McWilliams, D F; Ingham, S L; Doherty, M; Zhang, W

2017-03-20

Twenty-five percent of the British population over the age of 50 years experiences knee pain. Knee pain can limit physical ability and cause distress and bears significant socioeconomic costs. The objectives of this study were to develop and validate the first risk prediction model for incident knee pain in the Nottingham community and validate this internally within the Nottingham cohort and externally within the Osteoarthritis Initiative (OAI) cohort. A total of 1822 participants from the Nottingham community who were at risk for knee pain were followed for 12 years. Of this cohort, two-thirds (n = 1203) were used to develop the risk prediction model, and one-third (n = 619) were used to validate the model. Incident knee pain was defined as pain on most days for at least 1 month in the past 12 months. Predictors were age, sex, body mass index, pain elsewhere, prior knee injury and knee alignment. A Bayesian logistic regression model was used to determine the probability of an OR >1. The Hosmer-Lemeshow χ 2 statistic (HLS) was used for calibration, and ROC curve analysis was used for discrimination. The OAI cohort from the United States was also used to examine the performance of the model. A risk prediction model for knee pain incidence was developed using a Bayesian approach. The model had good calibration, with an HLS of 7.17 (p = 0.52) and moderate discriminative ability (ROC 0.70) in the community. Individual scenarios are given using the model. However, the model had poor calibration (HLS 5866.28, p < 0.01) and poor discriminative ability (ROC 0.54) in the OAI cohort. To our knowledge, this is the first risk prediction model for knee pain, regardless of underlying structural changes of knee osteoarthritis, in the community using a Bayesian modelling approach. The model appears to work well in a community-based population but not in individuals with a higher risk for knee osteoarthritis, and it may provide a convenient tool for use in primary care to predict the risk of knee pain in the general population.

Quantifying the uncertainty of nonpoint source attribution in distributed water quality models: A Bayesian assessment of SWAT's sediment export predictions

NASA Astrophysics Data System (ADS)

Wellen, Christopher; Arhonditsis, George B.; Long, Tanya; Boyd, Duncan

2014-11-01

Spatially distributed nonpoint source watershed models are essential tools to estimate the magnitude and sources of diffuse pollution. However, little work has been undertaken to understand the sources and ramifications of the uncertainty involved in their use. In this study we conduct the first Bayesian uncertainty analysis of the water quality components of the SWAT model, one of the most commonly used distributed nonpoint source models. Working in Southern Ontario, we apply three Bayesian configurations for calibrating SWAT to Redhill Creek, an urban catchment, and Grindstone Creek, an agricultural one. We answer four interrelated questions: can SWAT determine suspended sediment sources with confidence when end of basin data is used for calibration? How does uncertainty propagate from the discharge submodel to the suspended sediment submodels? Do the estimated sediment sources vary when different calibration approaches are used? Can we combine the knowledge gained from different calibration approaches? We show that: (i) despite reasonable fit at the basin outlet, the simulated sediment sources are subject to uncertainty sufficient to undermine the typical approach of reliance on a single, best fit simulation; (ii) more than a third of the uncertainty of sediment load predictions may stem from the discharge submodel; (iii) estimated sediment sources do vary significantly across the three statistical configurations of model calibration despite end-of-basin predictions being virtually identical; and (iv) Bayesian model averaging is an approach that can synthesize predictions when a number of adequate distributed models make divergent source apportionments. We conclude with recommendations for future research to reduce the uncertainty encountered when using distributed nonpoint source models for source apportionment.

Bayesian data analysis in population ecology: motivations, methods, and benefits

USGS Publications Warehouse

Dorazio, Robert

2016-01-01

During the 20th century ecologists largely relied on the frequentist system of inference for the analysis of their data. However, in the past few decades ecologists have become increasingly interested in the use of Bayesian methods of data analysis. In this article I provide guidance to ecologists who would like to decide whether Bayesian methods can be used to improve their conclusions and predictions. I begin by providing a concise summary of Bayesian methods of analysis, including a comparison of differences between Bayesian and frequentist approaches to inference when using hierarchical models. Next I provide a list of problems where Bayesian methods of analysis may arguably be preferred over frequentist methods. These problems are usually encountered in analyses based on hierarchical models of data. I describe the essentials required for applying modern methods of Bayesian computation, and I use real-world examples to illustrate these methods. I conclude by summarizing what I perceive to be the main strengths and weaknesses of using Bayesian methods to solve ecological inference problems.

Bayesian prediction of placebo analgesia in an instrumental learning model

PubMed Central

Jung, Won-Mo; Lee, Ye-Seul; Wallraven, Christian; Chae, Younbyoung

2017-01-01

Placebo analgesia can be primarily explained by the Pavlovian conditioning paradigm in which a passively applied cue becomes associated with less pain. In contrast, instrumental conditioning employs an active paradigm that might be more similar to clinical settings. In the present study, an instrumental conditioning paradigm involving a modified trust game in a simulated clinical situation was used to induce placebo analgesia. Additionally, Bayesian modeling was applied to predict the placebo responses of individuals based on their choices. Twenty-four participants engaged in a medical trust game in which decisions to receive treatment from either a doctor (more effective with high cost) or a pharmacy (less effective with low cost) were made after receiving a reference pain stimulus. In the conditioning session, the participants received lower levels of pain following both choices, while high pain stimuli were administered in the test session even after making the decision. The choice-dependent pain in the conditioning session was modulated in terms of both intensity and uncertainty. Participants reported significantly less pain when they chose the doctor or the pharmacy for treatment compared to the control trials. The predicted pain ratings based on Bayesian modeling showed significant correlations with the actual reports from participants for both of the choice categories. The instrumental conditioning paradigm allowed for the active choice of optional cues and was able to induce the placebo analgesia effect. Additionally, Bayesian modeling successfully predicted pain ratings in a simulated clinical situation that fits well with placebo analgesia induced by instrumental conditioning. PMID:28225816

The development of a probabilistic approach to forecast coastal change

USGS Publications Warehouse

Lentz, Erika E.; Hapke, Cheryl J.; Rosati, Julie D.; Wang, Ping; Roberts, Tiffany M.

2011-01-01

This study demonstrates the applicability of a Bayesian probabilistic model as an effective tool in predicting post-storm beach changes along sandy coastlines. Volume change and net shoreline movement are modeled for two study sites at Fire Island, New York in response to two extratropical storms in 2007 and 2009. Both study areas include modified areas adjacent to unmodified areas in morphologically different segments of coast. Predicted outcomes are evaluated against observed changes to test model accuracy and uncertainty along 163 cross-shore transects. Results show strong agreement in the cross validation of predictions vs. observations, with 70-82% accuracies reported. Although no consistent spatial pattern in inaccurate predictions could be determined, the highest prediction uncertainties appeared in locations that had been recently replenished. Further testing and model refinement are needed; however, these initial results show that Bayesian networks have the potential to serve as important decision-support tools in forecasting coastal change.

Prediction-error variance in Bayesian model updating: a comparative study

NASA Astrophysics Data System (ADS)

Asadollahi, Parisa; Li, Jian; Huang, Yong

2017-04-01

In Bayesian model updating, the likelihood function is commonly formulated by stochastic embedding in which the maximum information entropy probability model of prediction error variances plays an important role and it is Gaussian distribution subject to the first two moments as constraints. The selection of prediction error variances can be formulated as a model class selection problem, which automatically involves a trade-off between the average data-fit of the model class and the information it extracts from the data. Therefore, it is critical for the robustness in the updating of the structural model especially in the presence of modeling errors. To date, three ways of considering prediction error variances have been seem in the literature: 1) setting constant values empirically, 2) estimating them based on the goodness-of-fit of the measured data, and 3) updating them as uncertain parameters by applying Bayes' Theorem at the model class level. In this paper, the effect of different strategies to deal with the prediction error variances on the model updating performance is investigated explicitly. A six-story shear building model with six uncertain stiffness parameters is employed as an illustrative example. Transitional Markov Chain Monte Carlo is used to draw samples of the posterior probability density function of the structure model parameters as well as the uncertain prediction variances. The different levels of modeling uncertainty and complexity are modeled through three FE models, including a true model, a model with more complexity, and a model with modeling error. Bayesian updating is performed for the three FE models considering the three aforementioned treatments of the prediction error variances. The effect of number of measurements on the model updating performance is also examined in the study. The results are compared based on model class assessment and indicate that updating the prediction error variances as uncertain parameters at the model class level produces more robust results especially when the number of measurement is small.

A hybrid expectation maximisation and MCMC sampling algorithm to implement Bayesian mixture model based genomic prediction and QTL mapping.

PubMed

Wang, Tingting; Chen, Yi-Ping Phoebe; Bowman, Phil J; Goddard, Michael E; Hayes, Ben J

2016-09-21

Bayesian mixture models in which the effects of SNP are assumed to come from normal distributions with different variances are attractive for simultaneous genomic prediction and QTL mapping. These models are usually implemented with Monte Carlo Markov Chain (MCMC) sampling, which requires long compute times with large genomic data sets. Here, we present an efficient approach (termed HyB_BR), which is a hybrid of an Expectation-Maximisation algorithm, followed by a limited number of MCMC without the requirement for burn-in. To test prediction accuracy from HyB_BR, dairy cattle and human disease trait data were used. In the dairy cattle data, there were four quantitative traits (milk volume, protein kg, fat% in milk and fertility) measured in 16,214 cattle from two breeds genotyped for 632,002 SNPs. Validation of genomic predictions was in a subset of cattle either from the reference set or in animals from a third breeds that were not in the reference set. In all cases, HyB_BR gave almost identical accuracies to Bayesian mixture models implemented with full MCMC, however computational time was reduced by up to 1/17 of that required by full MCMC. The SNPs with high posterior probability of a non-zero effect were also very similar between full MCMC and HyB_BR, with several known genes affecting milk production in this category, as well as some novel genes. HyB_BR was also applied to seven human diseases with 4890 individuals genotyped for around 300 K SNPs in a case/control design, from the Welcome Trust Case Control Consortium (WTCCC). In this data set, the results demonstrated again that HyB_BR performed as well as Bayesian mixture models with full MCMC for genomic predictions and genetic architecture inference while reducing the computational time from 45 h with full MCMC to 3 h with HyB_BR. The results for quantitative traits in cattle and disease in humans demonstrate that HyB_BR can perform equally well as Bayesian mixture models implemented with full MCMC in terms of prediction accuracy, but with up to 17 times faster than the full MCMC implementations. The HyB_BR algorithm makes simultaneous genomic prediction, QTL mapping and inference of genetic architecture feasible in large genomic data sets.

Bayesian Inference of High-Dimensional Dynamical Ocean Models

NASA Astrophysics Data System (ADS)

Lin, J.; Lermusiaux, P. F. J.; Lolla, S. V. T.; Gupta, A.; Haley, P. J., Jr.

2015-12-01

This presentation addresses a holistic set of challenges in high-dimension ocean Bayesian nonlinear estimation: i) predict the probability distribution functions (pdfs) of large nonlinear dynamical systems using stochastic partial differential equations (PDEs); ii) assimilate data using Bayes' law with these pdfs; iii) predict the future data that optimally reduce uncertainties; and (iv) rank the known and learn the new model formulations themselves. Overall, we allow the joint inference of the state, equations, geometry, boundary conditions and initial conditions of dynamical models. Examples are provided for time-dependent fluid and ocean flows, including cavity, double-gyre and Strait flows with jets and eddies. The Bayesian model inference, based on limited observations, is illustrated first by the estimation of obstacle shapes and positions in fluid flows. Next, the Bayesian inference of biogeochemical reaction equations and of their states and parameters is presented, illustrating how PDE-based machine learning can rigorously guide the selection and discovery of complex ecosystem models. Finally, the inference of multiscale bottom gravity current dynamics is illustrated, motivated in part by classic overflows and dense water formation sites and their relevance to climate monitoring and dynamics. This is joint work with our MSEAS group at MIT.

A Bayesian network model for predicting aquatic toxicity mode of action using two dimensional theoretical molecular descriptors-abstract

EPA Science Inventory

The mode of toxic action (MoA) has been recognized as a key determinant of chemical toxicity but MoA classification in aquatic toxicology has been limited. We developed a Bayesian network model to classify aquatic toxicity mode of action using a recently published dataset contain...

A Bayesian Model of the Memory Colour Effect.

PubMed

Witzel, Christoph; Olkkonen, Maria; Gegenfurtner, Karl R

2018-01-01

According to the memory colour effect, the colour of a colour-diagnostic object is not perceived independently of the object itself. Instead, it has been shown through an achromatic adjustment method that colour-diagnostic objects still appear slightly in their typical colour, even when they are colourimetrically grey. Bayesian models provide a promising approach to capture the effect of prior knowledge on colour perception and to link these effects to more general effects of cue integration. Here, we model memory colour effects using prior knowledge about typical colours as priors for the grey adjustments in a Bayesian model. This simple model does not involve any fitting of free parameters. The Bayesian model roughly captured the magnitude of the measured memory colour effect for photographs of objects. To some extent, the model predicted observed differences in memory colour effects across objects. The model could not account for the differences in memory colour effects across different levels of realism in the object images. The Bayesian model provides a particularly simple account of memory colour effects, capturing some of the multiple sources of variation of these effects.

A Bayesian Model of the Memory Colour Effect

PubMed Central

Olkkonen, Maria; Gegenfurtner, Karl R.

2018-01-01

According to the memory colour effect, the colour of a colour-diagnostic object is not perceived independently of the object itself. Instead, it has been shown through an achromatic adjustment method that colour-diagnostic objects still appear slightly in their typical colour, even when they are colourimetrically grey. Bayesian models provide a promising approach to capture the effect of prior knowledge on colour perception and to link these effects to more general effects of cue integration. Here, we model memory colour effects using prior knowledge about typical colours as priors for the grey adjustments in a Bayesian model. This simple model does not involve any fitting of free parameters. The Bayesian model roughly captured the magnitude of the measured memory colour effect for photographs of objects. To some extent, the model predicted observed differences in memory colour effects across objects. The model could not account for the differences in memory colour effects across different levels of realism in the object images. The Bayesian model provides a particularly simple account of memory colour effects, capturing some of the multiple sources of variation of these effects. PMID:29760874

Bayesian methods in reliability

NASA Astrophysics Data System (ADS)

Sander, P.; Badoux, R.

1991-11-01

The present proceedings from a course on Bayesian methods in reliability encompasses Bayesian statistical methods and their computational implementation, models for analyzing censored data from nonrepairable systems, the traits of repairable systems and growth models, the use of expert judgment, and a review of the problem of forecasting software reliability. Specific issues addressed include the use of Bayesian methods to estimate the leak rate of a gas pipeline, approximate analyses under great prior uncertainty, reliability estimation techniques, and a nonhomogeneous Poisson process. Also addressed are the calibration sets and seed variables of expert judgment systems for risk assessment, experimental illustrations of the use of expert judgment for reliability testing, and analyses of the predictive quality of software-reliability growth models such as the Weibull order statistics.

A menu-driven software package of Bayesian nonparametric (and parametric) mixed models for regression analysis and density estimation.

PubMed

Karabatsos, George

2017-02-01

Most of applied statistics involves regression analysis of data. In practice, it is important to specify a regression model that has minimal assumptions which are not violated by data, to ensure that statistical inferences from the model are informative and not misleading. This paper presents a stand-alone and menu-driven software package, Bayesian Regression: Nonparametric and Parametric Models, constructed from MATLAB Compiler. Currently, this package gives the user a choice from 83 Bayesian models for data analysis. They include 47 Bayesian nonparametric (BNP) infinite-mixture regression models; 5 BNP infinite-mixture models for density estimation; and 31 normal random effects models (HLMs), including normal linear models. Each of the 78 regression models handles either a continuous, binary, or ordinal dependent variable, and can handle multi-level (grouped) data. All 83 Bayesian models can handle the analysis of weighted observations (e.g., for meta-analysis), and the analysis of left-censored, right-censored, and/or interval-censored data. Each BNP infinite-mixture model has a mixture distribution assigned one of various BNP prior distributions, including priors defined by either the Dirichlet process, Pitman-Yor process (including the normalized stable process), beta (two-parameter) process, normalized inverse-Gaussian process, geometric weights prior, dependent Dirichlet process, or the dependent infinite-probits prior. The software user can mouse-click to select a Bayesian model and perform data analysis via Markov chain Monte Carlo (MCMC) sampling. After the sampling completes, the software automatically opens text output that reports MCMC-based estimates of the model's posterior distribution and model predictive fit to the data. Additional text and/or graphical output can be generated by mouse-clicking other menu options. This includes output of MCMC convergence analyses, and estimates of the model's posterior predictive distribution, for selected functionals and values of covariates. The software is illustrated through the BNP regression analysis of real data.

Additive Genetic Variability and the Bayesian Alphabet

PubMed Central

Gianola, Daniel; de los Campos, Gustavo; Hill, William G.; Manfredi, Eduardo; Fernando, Rohan

2009-01-01

The use of all available molecular markers in statistical models for prediction of quantitative traits has led to what could be termed a genomic-assisted selection paradigm in animal and plant breeding. This article provides a critical review of some theoretical and statistical concepts in the context of genomic-assisted genetic evaluation of animals and crops. First, relationships between the (Bayesian) variance of marker effects in some regression models and additive genetic variance are examined under standard assumptions. Second, the connection between marker genotypes and resemblance between relatives is explored, and linkages between a marker-based model and the infinitesimal model are reviewed. Third, issues associated with the use of Bayesian models for marker-assisted selection, with a focus on the role of the priors, are examined from a theoretical angle. The sensitivity of a Bayesian specification that has been proposed (called “Bayes A”) with respect to priors is illustrated with a simulation. Methods that can solve potential shortcomings of some of these Bayesian regression procedures are discussed briefly. PMID:19620397

A Bayesian Hierarchical Model for Glacial Dynamics Based on the Shallow Ice Approximation and its Evaluation Using Analytical Solutions

NASA Astrophysics Data System (ADS)

Gopalan, Giri; Hrafnkelsson, Birgir; Aðalgeirsdóttir, Guðfinna; Jarosch, Alexander H.; Pálsson, Finnur

2018-03-01

Bayesian hierarchical modeling can assist the study of glacial dynamics and ice flow properties. This approach will allow glaciologists to make fully probabilistic predictions for the thickness of a glacier at unobserved spatio-temporal coordinates, and it will also allow for the derivation of posterior probability distributions for key physical parameters such as ice viscosity and basal sliding. The goal of this paper is to develop a proof of concept for a Bayesian hierarchical model constructed, which uses exact analytical solutions for the shallow ice approximation (SIA) introduced by Bueler et al. (2005). A suite of test simulations utilizing these exact solutions suggests that this approach is able to adequately model numerical errors and produce useful physical parameter posterior distributions and predictions. A byproduct of the development of the Bayesian hierarchical model is the derivation of a novel finite difference method for solving the SIA partial differential equation (PDE). An additional novelty of this work is the correction of numerical errors induced through a numerical solution using a statistical model. This error correcting process models numerical errors that accumulate forward in time and spatial variation of numerical errors between the dome, interior, and margin of a glacier.

Enhancing Flood Prediction Reliability Using Bayesian Model Averaging

NASA Astrophysics Data System (ADS)

Liu, Z.; Merwade, V.

2017-12-01

Uncertainty analysis is an indispensable part of modeling the hydrology and hydrodynamics of non-idealized environmental systems. Compared to reliance on prediction from one model simulation, using on ensemble of predictions that consider uncertainty from different sources is more reliable. In this study, Bayesian model averaging (BMA) is applied to Black River watershed in Arkansas and Missouri by combining multi-model simulations to get reliable deterministic water stage and probabilistic inundation extent predictions. The simulation ensemble is generated from 81 LISFLOOD-FP subgrid model configurations that include uncertainty from channel shape, channel width, channel roughness and discharge. Model simulation outputs are trained with observed water stage data during one flood event, and BMA prediction ability is validated for another flood event. Results from this study indicate that BMA does not always outperform all members in the ensemble, but it provides relatively robust deterministic flood stage predictions across the basin. Station based BMA (BMA_S) water stage prediction has better performance than global based BMA (BMA_G) prediction which is superior to the ensemble mean prediction. Additionally, high-frequency flood inundation extent (probability greater than 60%) in BMA_G probabilistic map is more accurate than the probabilistic flood inundation extent based on equal weights.

Bayesian analysis of rare events

DOE Office of Scientific and Technical Information (OSTI.GOV)

Straub, Daniel, E-mail: straub@tum.de; Papaioannou, Iason; Betz, Wolfgang

2016-06-01

In many areas of engineering and science there is an interest in predicting the probability of rare events, in particular in applications related to safety and security. Increasingly, such predictions are made through computer models of physical systems in an uncertainty quantification framework. Additionally, with advances in IT, monitoring and sensor technology, an increasing amount of data on the performance of the systems is collected. This data can be used to reduce uncertainty, improve the probability estimates and consequently enhance the management of rare events and associated risks. Bayesian analysis is the ideal method to include the data into themore » probabilistic model. It ensures a consistent probabilistic treatment of uncertainty, which is central in the prediction of rare events, where extrapolation from the domain of observation is common. We present a framework for performing Bayesian updating of rare event probabilities, termed BUS. It is based on a reinterpretation of the classical rejection-sampling approach to Bayesian analysis, which enables the use of established methods for estimating probabilities of rare events. By drawing upon these methods, the framework makes use of their computational efficiency. These methods include the First-Order Reliability Method (FORM), tailored importance sampling (IS) methods and Subset Simulation (SuS). In this contribution, we briefly review these methods in the context of the BUS framework and investigate their applicability to Bayesian analysis of rare events in different settings. We find that, for some applications, FORM can be highly efficient and is surprisingly accurate, enabling Bayesian analysis of rare events with just a few model evaluations. In a general setting, BUS implemented through IS and SuS is more robust and flexible.« less

How Reliable is Bayesian Model Averaging Under Noisy Data? Statistical Assessment and Implications for Robust Model Selection

NASA Astrophysics Data System (ADS)

Schöniger, Anneli; Wöhling, Thomas; Nowak, Wolfgang

2014-05-01

Bayesian model averaging ranks the predictive capabilities of alternative conceptual models based on Bayes' theorem. The individual models are weighted with their posterior probability to be the best one in the considered set of models. Finally, their predictions are combined into a robust weighted average and the predictive uncertainty can be quantified. This rigorous procedure does, however, not yet account for possible instabilities due to measurement noise in the calibration data set. This is a major drawback, since posterior model weights may suffer a lack of robustness related to the uncertainty in noisy data, which may compromise the reliability of model ranking. We present a new statistical concept to account for measurement noise as source of uncertainty for the weights in Bayesian model averaging. Our suggested upgrade reflects the limited information content of data for the purpose of model selection. It allows us to assess the significance of the determined posterior model weights, the confidence in model selection, and the accuracy of the quantified predictive uncertainty. Our approach rests on a brute-force Monte Carlo framework. We determine the robustness of model weights against measurement noise by repeatedly perturbing the observed data with random realizations of measurement error. Then, we analyze the induced variability in posterior model weights and introduce this "weighting variance" as an additional term into the overall prediction uncertainty analysis scheme. We further determine the theoretical upper limit in performance of the model set which is imposed by measurement noise. As an extension to the merely relative model ranking, this analysis provides a measure of absolute model performance. To finally decide, whether better data or longer time series are needed to ensure a robust basis for model selection, we resample the measurement time series and assess the convergence of model weights for increasing time series length. We illustrate our suggested approach with an application to model selection between different soil-plant models following up on a study by Wöhling et al. (2013). Results show that measurement noise compromises the reliability of model ranking and causes a significant amount of weighting uncertainty, if the calibration data time series is not long enough to compensate for its noisiness. This additional contribution to the overall predictive uncertainty is neglected without our approach. Thus, we strongly advertise to include our suggested upgrade in the Bayesian model averaging routine.

A Bayesian modification to the Jelinski-Moranda software reliability growth model

NASA Technical Reports Server (NTRS)

Littlewood, B.; Sofer, A.

1983-01-01

The Jelinski-Moranda (JM) model for software reliability was examined. It is suggested that a major reason for the poor results given by this model is the poor performance of the maximum likelihood method (ML) of parameter estimation. A reparameterization and Bayesian analysis, involving a slight modelling change, are proposed. It is shown that this new Bayesian-Jelinski-Moranda model (BJM) is mathematically quite tractable, and several metrics of interest to practitioners are obtained. The BJM and JM models are compared by using several sets of real software failure data collected and in all cases the BJM model gives superior reliability predictions. A change in the assumption which underlay both models to present the debugging process more accurately is discussed.

Developing and Testing a Model to Predict Outcomes of Organizational Change

PubMed Central

Gustafson, David H; Sainfort, François; Eichler, Mary; Adams, Laura; Bisognano, Maureen; Steudel, Harold

2003-01-01

Objective To test the effectiveness of a Bayesian model employing subjective probability estimates for predicting success and failure of health care improvement projects. Data Sources Experts' subjective assessment data for model development and independent retrospective data on 221 healthcare improvement projects in the United States, Canada, and the Netherlands collected between 1996 and 2000 for validation. Methods A panel of theoretical and practical experts and literature in organizational change were used to identify factors predicting the outcome of improvement efforts. A Bayesian model was developed to estimate probability of successful change using subjective estimates of likelihood ratios and prior odds elicited from the panel of experts. A subsequent retrospective empirical analysis of change efforts in 198 health care organizations was performed to validate the model. Logistic regression and ROC analysis were used to evaluate the model's performance using three alternative definitions of success. Data Collection For the model development, experts' subjective assessments were elicited using an integrative group process. For the validation study, a staff person intimately involved in each improvement project responded to a written survey asking questions about model factors and project outcomes. Results Logistic regression chi-square statistics and areas under the ROC curve demonstrated a high level of model performance in predicting success. Chi-square statistics were significant at the 0.001 level and areas under the ROC curve were greater than 0.84. Conclusions A subjective Bayesian model was effective in predicting the outcome of actual improvement projects. Additional prospective evaluations as well as testing the impact of this model as an intervention are warranted. PMID:12785571

Bayesian Nonparametric Prediction and Statistical Inference

DTIC Science & Technology

1989-09-07

Kadane, J. (1980), "Bayesian decision theory and the sim- plification of models," in Evaluation of Econometric Models, J. Kmenta and J. Ramsey , eds...the random model and weighted least squares regression," in Evaluation of Econometric Models, ed. by J. Kmenta and J. Ramsey , Academic Press, 197-217...likelihood function. On the other hand, H. Jeffreys’s theory of hypothesis testing covers the most important situations in which the prior is not diffuse. See

Using discharge data to reduce structural deficits in a hydrological model with a Bayesian inference approach and the implications for the prediction of critical source areas

NASA Astrophysics Data System (ADS)

Frey, M. P.; Stamm, C.; Schneider, M. K.; Reichert, P.

2011-12-01

A distributed hydrological model was used to simulate the distribution of fast runoff formation as a proxy for critical source areas for herbicide pollution in a small agricultural catchment in Switzerland. We tested to what degree predictions based on prior knowledge without local measurements could be improved upon relying on observed discharge. This learning process consisted of five steps: For the prior prediction (step 1), knowledge of the model parameters was coarse and predictions were fairly uncertain. In the second step, discharge data were used to update the prior parameter distribution. Effects of uncertainty in input data and model structure were accounted for by an autoregressive error model. This step decreased the width of the marginal distributions of parameters describing the lower boundary (percolation rates) but hardly affected soil hydraulic parameters. Residual analysis (step 3) revealed model structure deficits. We modified the model, and in the subsequent Bayesian updating (step 4) the widths of the posterior marginal distributions were reduced for most parameters compared to those of the prior. This incremental procedure led to a strong reduction in the uncertainty of the spatial prediction. Thus, despite only using spatially integrated data (discharge), the spatially distributed effect of the improved model structure can be expected to improve the spatially distributed predictions also. The fifth step consisted of a test with independent spatial data on herbicide losses and revealed ambiguous results. The comparison depended critically on the ratio of event to preevent water that was discharged. This ratio cannot be estimated from hydrological data only. The results demonstrate that the value of local data is strongly dependent on a correct model structure. An iterative procedure of Bayesian updating, model testing, and model modification is suggested.

Assessment of Manual Operation Time for the Manufacturing of Thin Film Transistor Liquid Crystal Display: A Bayesian Approach

NASA Astrophysics Data System (ADS)

Shen, Chien-wen

2009-01-01

During the processes of TFT-LCD manufacturing, steps like visual inspection of panel surface defects still heavily rely on manual operations. As the manual inspection time of TFT-LCD manufacturing could range from 4 hours to 1 day, the reliability of time forecasting is thus important for production planning, scheduling and customer response. This study would like to propose a practical and easy-to-implement prediction model through the approach of Bayesian networks for time estimation of manual operated procedures in TFT-LCD manufacturing. Given the lack of prior knowledge about manual operation time, algorithms of necessary path condition and expectation-maximization are used for structural learning and estimation of conditional probability distributions respectively. This study also applied Bayesian inference to evaluate the relationships between explanatory variables and manual operation time. With the empirical applications of this proposed forecasting model, approach of Bayesian networks demonstrates its practicability and prediction accountability.

K-ε Turbulence Model Parameter Estimates Using an Approximate Self-similar Jet-in-Crossflow Solution

DOE PAGES

DeChant, Lawrence; Ray, Jaideep; Lefantzi, Sophia; ...

2017-06-09

The k-ε turbulence model has been described as perhaps “the most widely used complete turbulence model.” This family of heuristic Reynolds Averaged Navier-Stokes (RANS) turbulence closures is supported by a suite of model parameters that have been estimated by demanding the satisfaction of well-established canonical flows such as homogeneous shear flow, log-law behavior, etc. While this procedure does yield a set of so-called nominal parameters, it is abundantly clear that they do not provide a universally satisfactory turbulence model that is capable of simulating complex flows. Recent work on the Bayesian calibration of the k-ε model using jet-in-crossflow wind tunnelmore » data has yielded parameter estimates that are far more predictive than nominal parameter values. In this paper, we develop a self-similar asymptotic solution for axisymmetric jet-in-crossflow interactions and derive analytical estimates of the parameters that were inferred using Bayesian calibration. The self-similar method utilizes a near field approach to estimate the turbulence model parameters while retaining the classical far-field scaling to model flow field quantities. Our parameter values are seen to be far more predictive than the nominal values, as checked using RANS simulations and experimental measurements. They are also closer to the Bayesian estimates than the nominal parameters. A traditional simplified jet trajectory model is explicitly related to the turbulence model parameters and is shown to yield good agreement with measurement when utilizing the analytical derived turbulence model coefficients. Finally, the close agreement between the turbulence model coefficients obtained via Bayesian calibration and the analytically estimated coefficients derived in this paper is consistent with the contention that the Bayesian calibration approach is firmly rooted in the underlying physical description.« less

Maximum a posteriori Bayesian estimation of mycophenolic Acid area under the concentration-time curve: is this clinically useful for dosage prediction yet?

PubMed

Staatz, Christine E; Tett, Susan E

2011-12-01

This review seeks to summarize the available data about Bayesian estimation of area under the plasma concentration-time curve (AUC) and dosage prediction for mycophenolic acid (MPA) and evaluate whether sufficient evidence is available for routine use of Bayesian dosage prediction in clinical practice. A literature search identified 14 studies that assessed the predictive performance of maximum a posteriori Bayesian estimation of MPA AUC and one report that retrospectively evaluated how closely dosage recommendations based on Bayesian forecasting achieved targeted MPA exposure. Studies to date have mostly been undertaken in renal transplant recipients, with limited investigation in patients treated with MPA for autoimmune disease or haematopoietic stem cell transplantation. All of these studies have involved use of the mycophenolate mofetil (MMF) formulation of MPA, rather than the enteric-coated mycophenolate sodium (EC-MPS) formulation. Bias associated with estimation of MPA AUC using Bayesian forecasting was generally less than 10%. However some difficulties with imprecision was evident, with values ranging from 4% to 34% (based on estimation involving two or more concentration measurements). Evaluation of whether MPA dosing decisions based on Bayesian forecasting (by the free website service https://pharmaco.chu-limoges.fr) achieved target drug exposure has only been undertaken once. When MMF dosage recommendations were applied by clinicians, a higher proportion (72-80%) of subsequent estimated MPA AUC values were within the 30-60 mg · h/L target range, compared with when dosage recommendations were not followed (only 39-57% within target range). Such findings provide evidence that Bayesian dosage prediction is clinically useful for achieving target MPA AUC. This study, however, was retrospective and focussed only on adult renal transplant recipients. Furthermore, in this study, Bayesian-generated AUC estimations and dosage predictions were not compared with a later full measured AUC but rather with a further AUC estimate based on a second Bayesian analysis. This study also provided some evidence that a useful monitoring schedule for MPA AUC following adult renal transplant would be every 2 weeks during the first month post-transplant, every 1-3 months between months 1 and 12, and each year thereafter. It will be interesting to see further validations in different patient groups using the free website service. In summary, the predictive performance of Bayesian estimation of MPA, comparing estimated with measured AUC values, has been reported in several studies. However, the next step of predicting dosages based on these Bayesian-estimated AUCs, and prospectively determining how closely these predicted dosages give drug exposure matching targeted AUCs, remains largely unaddressed. Further prospective studies are required, particularly in non-renal transplant patients and with the EC-MPS formulation. Other important questions remain to be answered, such as: do Bayesian forecasting methods devised to date use the best population pharmacokinetic models or most accurate algorithms; are the methods simple to use for routine clinical practice; do the algorithms actually improve dosage estimations beyond empirical recommendations in all groups that receive MPA therapy; and, importantly, do the dosage predictions, when followed, improve patient health outcomes?

Bayesian truncation errors in chiral effective field theory: model checking and accounting for correlations

NASA Astrophysics Data System (ADS)

Melendez, Jordan; Wesolowski, Sarah; Furnstahl, Dick

2017-09-01

Chiral effective field theory (EFT) predictions are necessarily truncated at some order in the EFT expansion, which induces an error that must be quantified for robust statistical comparisons to experiment. A Bayesian model yields posterior probability distribution functions for these errors based on expectations of naturalness encoded in Bayesian priors and the observed order-by-order convergence pattern of the EFT. As a general example of a statistical approach to truncation errors, the model was applied to chiral EFT for neutron-proton scattering using various semi-local potentials of Epelbaum, Krebs, and Meißner (EKM). Here we discuss how our model can learn correlation information from the data and how to perform Bayesian model checking to validate that the EFT is working as advertised. Supported in part by NSF PHY-1614460 and DOE NUCLEI SciDAC DE-SC0008533.

Technical Topic 3.2.2.d Bayesian and Non-Parametric Statistics: Integration of Neural Networks with Bayesian Networks for Data Fusion and Predictive Modeling

DTIC Science & Technology

2016-05-31

and included explosives such as TATP, HMTD, RDX, RDX, ammonium nitrate , potassium perchlorate, potassium nitrate , sugar, and TNT. The approach...Distribution Unlimited UU UU UU UU 31-05-2016 15-Apr-2014 14-Jan-2015 Final Report: Technical Topic 3.2.2. d Bayesian and Non- parametric Statistics...of Papers published in non peer-reviewed journals: Final Report: Technical Topic 3.2.2. d Bayesian and Non-parametric Statistics: Integration of Neural

Context Relevant Prediction Model for COPD Domain Using Bayesian Belief Network

PubMed Central

Saleh, Lokman; Ajami, Hicham; Mili, Hafedh

2017-01-01

In the last three decades, researchers have examined extensively how context-aware systems can assist people, specifically those suffering from incurable diseases, to help them cope with their medical illness. Over the years, a huge number of studies on Chronic Obstructive Pulmonary Disease (COPD) have been published. However, how to derive relevant attributes and early detection of COPD exacerbations remains a challenge. In this research work, we will use an efficient algorithm to select relevant attributes where there is no proper approach in this domain. Such algorithm predicts exacerbations with high accuracy by adding discretization process, and organizes the pertinent attributes in priority order based on their impact to facilitate the emergency medical treatment. In this paper, we propose an extension of our existing Helper Context-Aware Engine System (HCES) for COPD. This project uses Bayesian network algorithm to depict the dependency between the COPD symptoms (attributes) in order to overcome the insufficiency and the independency hypothesis of naïve Bayesian. In addition, the dependency in Bayesian network is realized using TAN algorithm rather than consulting pneumologists. All these combined algorithms (discretization, selection, dependency, and the ordering of the relevant attributes) constitute an effective prediction model, comparing to effective ones. Moreover, an investigation and comparison of different scenarios of these algorithms are also done to verify which sequence of steps of prediction model gives more accurate results. Finally, we designed and validated a computer-aided support application to integrate different steps of this model. The findings of our system HCES has shown promising results using Area Under Receiver Operating Characteristic (AUC = 81.5%). PMID:28644419

Genomic Bayesian functional regression models with interactions for predicting wheat grain yield using hyper-spectral image data.

PubMed

Montesinos-López, Abelardo; Montesinos-López, Osval A; Cuevas, Jaime; Mata-López, Walter A; Burgueño, Juan; Mondal, Sushismita; Huerta, Julio; Singh, Ravi; Autrique, Enrique; González-Pérez, Lorena; Crossa, José

2017-01-01

Modern agriculture uses hyperspectral cameras that provide hundreds of reflectance data at discrete narrow bands in many environments. These bands often cover the whole visible light spectrum and part of the infrared and ultraviolet light spectra. With the bands, vegetation indices are constructed for predicting agronomically important traits such as grain yield and biomass. However, since vegetation indices only use some wavelengths (referred to as bands), we propose using all bands simultaneously as predictor variables for the primary trait grain yield; results of several multi-environment maize (Aguate et al. in Crop Sci 57(5):1-8, 2017) and wheat (Montesinos-López et al. in Plant Methods 13(4):1-23, 2017) breeding trials indicated that using all bands produced better prediction accuracy than vegetation indices. However, until now, these prediction models have not accounted for the effects of genotype × environment (G × E) and band × environment (B × E) interactions incorporating genomic or pedigree information. In this study, we propose Bayesian functional regression models that take into account all available bands, genomic or pedigree information, the main effects of lines and environments, as well as G × E and B × E interaction effects. The data set used is comprised of 976 wheat lines evaluated for grain yield in three environments (Drought, Irrigated and Reduced Irrigation). The reflectance data were measured in 250 discrete narrow bands ranging from 392 to 851 nm (nm). The proposed Bayesian functional regression models were implemented using two types of basis: B-splines and Fourier. Results of the proposed Bayesian functional regression models, including all the wavelengths for predicting grain yield, were compared with results from conventional models with and without bands. We observed that the models with B × E interaction terms were the most accurate models, whereas the functional regression models (with B-splines and Fourier basis) and the conventional models performed similarly in terms of prediction accuracy. However, the functional regression models are more parsimonious and computationally more efficient because the number of beta coefficients to be estimated is 21 (number of basis), rather than estimating the 250 regression coefficients for all bands. In this study adding pedigree or genomic information did not increase prediction accuracy.

Predicting Mycobacterium tuberculosis Complex Clades Using Knowledge-Based Bayesian Networks

PubMed Central

Bennett, Kristin P.

2014-01-01

We develop a novel approach for incorporating expert rules into Bayesian networks for classification of Mycobacterium tuberculosis complex (MTBC) clades. The proposed knowledge-based Bayesian network (KBBN) treats sets of expert rules as prior distributions on the classes. Unlike prior knowledge-based support vector machine approaches which require rules expressed as polyhedral sets, KBBN directly incorporates the rules without any modification. KBBN uses data to refine rule-based classifiers when the rule set is incomplete or ambiguous. We develop a predictive KBBN model for 69 MTBC clades found in the SITVIT international collection. We validate the approach using two testbeds that model knowledge of the MTBC obtained from two different experts and large DNA fingerprint databases to predict MTBC genetic clades and sublineages. These models represent strains of MTBC using high-throughput biomarkers called spacer oligonucleotide types (spoligotypes), since these are routinely gathered from MTBC isolates of tuberculosis (TB) patients. Results show that incorporating rules into problems can drastically increase classification accuracy if data alone are insufficient. The SITVIT KBBN is publicly available for use on the World Wide Web. PMID:24864238

Preferential sampling and Bayesian geostatistics: Statistical modeling and examples.

PubMed

Cecconi, Lorenzo; Grisotto, Laura; Catelan, Dolores; Lagazio, Corrado; Berrocal, Veronica; Biggeri, Annibale

2016-08-01

Preferential sampling refers to any situation in which the spatial process and the sampling locations are not stochastically independent. In this paper, we present two examples of geostatistical analysis in which the usual assumption of stochastic independence between the point process and the measurement process is violated. To account for preferential sampling, we specify a flexible and general Bayesian geostatistical model that includes a shared spatial random component. We apply the proposed model to two different case studies that allow us to highlight three different modeling and inferential aspects of geostatistical modeling under preferential sampling: (1) continuous or finite spatial sampling frame; (2) underlying causal model and relevant covariates; and (3) inferential goals related to mean prediction surface or prediction uncertainty. © The Author(s) 2016.

Bayesian Mapping Reveals That Attention Boosts Neural Responses to Predicted and Unpredicted Stimuli.

PubMed

Garrido, Marta I; Rowe, Elise G; Halász, Veronika; Mattingley, Jason B

2018-05-01

Predictive coding posits that the human brain continually monitors the environment for regularities and detects inconsistencies. It is unclear, however, what effect attention has on expectation processes, as there have been relatively few studies and the results of these have yielded contradictory findings. Here, we employed Bayesian model comparison to adjudicate between 2 alternative computational models. The "Opposition" model states that attention boosts neural responses equally to predicted and unpredicted stimuli, whereas the "Interaction" model assumes that attentional boosting of neural signals depends on the level of predictability. We designed a novel, audiospatial attention task that orthogonally manipulated attention and prediction by playing oddball sequences in either the attended or unattended ear. We observed sensory prediction error responses, with electroencephalography, across all attentional manipulations. Crucially, posterior probability maps revealed that, overall, the Opposition model better explained scalp and source data, suggesting that attention boosts responses to predicted and unpredicted stimuli equally. Furthermore, Dynamic Causal Modeling showed that these Opposition effects were expressed in plastic changes within the mismatch negativity network. Our findings provide empirical evidence for a computational model of the opposing interplay of attention and expectation in the brain.

Prognostics of lithium-ion batteries based on Dempster-Shafer theory and the Bayesian Monte Carlo method

NASA Astrophysics Data System (ADS)

He, Wei; Williard, Nicholas; Osterman, Michael; Pecht, Michael

A new method for state of health (SOH) and remaining useful life (RUL) estimations for lithium-ion batteries using Dempster-Shafer theory (DST) and the Bayesian Monte Carlo (BMC) method is proposed. In this work, an empirical model based on the physical degradation behavior of lithium-ion batteries is developed. Model parameters are initialized by combining sets of training data based on DST. BMC is then used to update the model parameters and predict the RUL based on available data through battery capacity monitoring. As more data become available, the accuracy of the model in predicting RUL improves. Two case studies demonstrating this approach are presented.

Predictors of Outcome in Traumatic Brain Injury: New Insight Using Receiver Operating Curve Indices and Bayesian Network Analysis.

PubMed

Zador, Zsolt; Sperrin, Matthew; King, Andrew T

2016-01-01

Traumatic brain injury remains a global health problem. Understanding the relative importance of outcome predictors helps optimize our treatment strategies by informing assessment protocols, clinical decisions and trial designs. In this study we establish importance ranking for outcome predictors based on receiver operating indices to identify key predictors of outcome and create simple predictive models. We then explore the associations between key outcome predictors using Bayesian networks to gain further insight into predictor importance. We analyzed the corticosteroid randomization after significant head injury (CRASH) trial database of 10008 patients and included patients for whom demographics, injury characteristics, computer tomography (CT) findings and Glasgow Outcome Scale (GCS) were recorded (total of 13 predictors, which would be available to clinicians within a few hours following the injury in 6945 patients). Predictions of clinical outcome (death or severe disability at 6 months) were performed using logistic regression models with 5-fold cross validation. Predictive performance was measured using standardized partial area (pAUC) under the receiver operating curve (ROC) and we used Delong test for comparisons. Variable importance ranking was based on pAUC targeted at specificity (pAUCSP) and sensitivity (pAUCSE) intervals of 90-100%. Probabilistic associations were depicted using Bayesian networks. Complete AUC analysis showed very good predictive power (AUC = 0.8237, 95% CI: 0.8138-0.8336) for the complete model. Specificity focused importance ranking highlighted age, pupillary, motor responses, obliteration of basal cisterns/3rd ventricle and midline shift. Interestingly when targeting model sensitivity, the highest-ranking variables were age, severe extracranial injury, verbal response, hematoma on CT and motor response. Simplified models, which included only these key predictors, had similar performance (pAUCSP = 0.6523, 95% CI: 0.6402-0.6641 and pAUCSE = 0.6332, 95% CI: 0.62-0.6477) compared to the complete models (pAUCSP = 0.6664, 95% CI: 0.6543-0.679, pAUCSE = 0.6436, 95% CI: 0.6289-0.6585, de Long p value 0.1165 and 0.3448 respectively). Bayesian networks showed the predictors that did not feature in the simplified models were associated with those that did. We demonstrate that importance based variable selection allows simplified predictive models to be created while maintaining prediction accuracy. Variable selection targeting specificity confirmed key components of clinical assessment in TBI whereas sensitivity based ranking suggested extracranial injury as one of the important predictors. These results help refine our approach to head injury assessment, decision-making and outcome prediction targeted at model sensitivity and specificity. Bayesian networks proved to be a comprehensive tool for depicting probabilistic associations for key predictors giving insight into why the simplified model has maintained accuracy.

Prediction of Sybil attack on WSN using Bayesian network and swarm intelligence

NASA Astrophysics Data System (ADS)

Muraleedharan, Rajani; Ye, Xiang; Osadciw, Lisa Ann

2008-04-01

Security in wireless sensor networks is typically sacrificed or kept minimal due to limited resources such as memory and battery power. Hence, the sensor nodes are prone to Denial-of-service attacks and detecting the threats is crucial in any application. In this paper, the Sybil attack is analyzed and a novel prediction method, combining Bayesian algorithm and Swarm Intelligence (SI) is proposed. Bayesian Networks (BN) is used in representing and reasoning problems, by modeling the elements of uncertainty. The decision from the BN is applied to SI forming an Hybrid Intelligence Scheme (HIS) to re-route the information and disconnecting the malicious nodes in future routes. A performance comparison based on the prediction using HIS vs. Ant System (AS) helps in prioritizing applications where decisions are time-critical.

A Bayesian network to predict vulnerability to sea-level rise: data report

USGS Publications Warehouse

Gutierrez, Benjamin T.; Plant, Nathaniel G.; Thieler, E. Robert

2011-01-01

During the 21st century, sea-level rise is projected to have a wide range of effects on coastal environments, development, and infrastructure. Consequently, there has been an increased focus on developing modeling or other analytical approaches to evaluate potential impacts to inform coastal management. This report provides the data that were used to develop and evaluate the performance of a Bayesian network designed to predict long-term shoreline change due to sea-level rise. The data include local rates of relative sea-level rise, wave height, tide range, geomorphic classification, coastal slope, and shoreline-change rate compiled as part of the U.S. Geological Survey Coastal Vulnerability Index for the U.S. Atlantic coast. In this project, the Bayesian network is used to define relationships among driving forces, geologic constraints, and coastal responses. Using this information, the Bayesian network is used to make probabilistic predictions of shoreline change in response to different future sea-level-rise scenarios.

Calculation of Crystallographic Texture of BCC Steels During Cold Rolling

NASA Astrophysics Data System (ADS)

Das, Arpan

2017-05-01

BCC alloys commonly tend to develop strong fibre textures and often represent as isointensity diagrams in φ 1 sections or by fibre diagrams. Alpha fibre in bcc steels is generally characterised by <110> crystallographic axis parallel to the rolling direction. The objective of present research is to correlate carbon content, carbide dispersion, rolling reduction, Euler angles (ϕ) (when φ 1 = 0° and φ 2 = 45° along alpha fibre) and the resulting alpha fibre texture orientation intensity. In the present research, Bayesian neural computation has been employed to correlate these and compare with the existing feed-forward neural network model comprehensively. Excellent match to the measured texture data within the bounding box of texture training data set has been already predicted through the feed-forward neural network model by other researchers. Feed-forward neural network prediction outside the bounds of training texture data showed deviations from the expected values. Currently, Bayesian computation has been similarly applied to confirm that the predictions are reasonable in the context of basic metallurgical principles, and matched better outside the bounds of training texture data set than the reported feed-forward neural network. Bayesian computation puts error bars on predicted values and allows significance of each individual parameters to be estimated. Additionally, it is also possible by Bayesian computation to estimate the isolated influence of particular variable such as carbon concentration, which exactly cannot in practice be varied independently. This shows the ability of the Bayesian neural network to examine the new phenomenon in situations where the data cannot be accessed through experiments.

Impact of Bias-Correction Type and Conditional Training on Bayesian Model Averaging over the Northeast United States

Treesearch

Michael J. Erickson; Brian A. Colle; Joseph J. Charney

2012-01-01

The performance of a multimodel ensemble over the northeast United States is evaluated before and after applying bias correction and Bayesian model averaging (BMA). The 13-member Stony Brook University (SBU) ensemble at 0000 UTC is combined with the 21-member National Centers for Environmental Prediction (NCEP) Short-Range Ensemble Forecast (SREF) system at 2100 UTC....

A bayesian hierarchical model for spatio-temporal prediction and uncertainty assessment using repeat LiDAR acquisitions for the Kenai Peninsula, AK, USA

Treesearch

Chad Babcock; Hans Andersen; Andrew O. Finley; Bruce D. Cook

2015-01-01

Models leveraging repeat LiDAR and field collection campaigns may be one possible mechanism to monitor carbon flux in remote forested regions. Here, we look to the spatio-temporally data-rich Kenai Peninsula in Alaska, USA to examine the potential for Bayesian spatio-temporal mapping of terrestrial forest carbon storage and uncertainty.

Multi-Scale Validation of a Nanodiamond Drug Delivery System and Multi-Scale Engineering Education

ERIC Educational Resources Information Center

Schwalbe, Michelle Kristin

2010-01-01

This dissertation has two primary concerns: (i) evaluating the uncertainty and prediction capabilities of a nanodiamond drug delivery model using Bayesian calibration and bias correction, and (ii) determining conceptual difficulties of multi-scale analysis from an engineering education perspective. A Bayesian uncertainty quantification scheme…

Bayesian-based estimation of acoustic surface impedance: Finite difference frequency domain approach.

PubMed

Bockman, Alexander; Fackler, Cameron; Xiang, Ning

2015-04-01

Acoustic performance for an interior requires an accurate description of the boundary materials' surface acoustic impedance. Analytical methods may be applied to a small class of test geometries, but inverse numerical methods provide greater flexibility. The parameter estimation problem requires minimizing prediction vice observed acoustic field pressure. The Bayesian-network sampling approach presented here mitigates other methods' susceptibility to noise inherent to the experiment, model, and numerics. A geometry agnostic method is developed here and its parameter estimation performance is demonstrated for an air-backed micro-perforated panel in an impedance tube. Good agreement is found with predictions from the ISO standard two-microphone, impedance-tube method, and a theoretical model for the material. Data by-products exclusive to a Bayesian approach are analyzed to assess sensitivity of the method to nuisance parameters.

Logistic Mixed Models to Investigate Implicit and Explicit Belief Tracking.

PubMed

Lages, Martin; Scheel, Anne

2016-01-01

We investigated the proposition of a two-systems Theory of Mind in adults' belief tracking. A sample of N = 45 participants predicted the choice of one of two opponent players after observing several rounds in an animated card game. Three matches of this card game were played and initial gaze direction on target and subsequent choice predictions were recorded for each belief task and participant. We conducted logistic regressions with mixed effects on the binary data and developed Bayesian logistic mixed models to infer implicit and explicit mentalizing in true belief and false belief tasks. Although logistic regressions with mixed effects predicted the data well a Bayesian logistic mixed model with latent task- and subject-specific parameters gave a better account of the data. As expected explicit choice predictions suggested a clear understanding of true and false beliefs (TB/FB). Surprisingly, however, model parameters for initial gaze direction also indicated belief tracking. We discuss why task-specific parameters for initial gaze directions are different from choice predictions yet reflect second-order perspective taking.

Static and transient performance prediction for CFB boilers using a Bayesian-Gaussian Neural Network

NASA Astrophysics Data System (ADS)

Ye, Haiwen; Ni, Weidou

1997-06-01

A Bayesian-Gaussian Neural Network (BGNN) is put forward in this paper to predict the static and transient performance of Circulating Fluidized Bed (CFB) boilers. The advantages of this network over Back-Propagation Neural Networks (BPNNs), easier determination of topology, simpler and time saving in training process as well as self-organizing ability, make this network more practical in on-line performance prediction for complicated processes. Simulation shows that this network is comparable to the BPNNs in predicting the performance of CFB boilers. Good and practical on-line performance predictions are essential for operation guide and model predictive control of CFB boilers, which are under research by the authors.

Prediction and repeatability of milk coagulation properties and curd-firming modeling parameters of ovine milk using Fourier-transform infrared spectroscopy and Bayesian models.

PubMed

Ferragina, A; Cipolat-Gotet, C; Cecchinato, A; Pazzola, M; Dettori, M L; Vacca, G M; Bittante, G

2017-05-01

The aim of this study was to apply Bayesian models to the Fourier-transform infrared spectroscopy spectra of individual sheep milk samples to derive calibration equations to predict traditional and modeled milk coagulation properties (MCP), and to assess the repeatability of MCP measures and their predictions. Data consisted of 1,002 individual milk samples collected from Sarda ewes reared in 22 farms in the region of Sardinia (Italy) for which MCP and modeled curd-firming parameters were available. Two milk samples were taken from 87 ewes and analyzed with the aim of estimating repeatability, whereas a single sample was taken from the other 915 ewes. Therefore, a total of 1,089 analyses were performed. For each sample, 2 spectra in the infrared region 5,011 to 925 cm -1 were available and averaged before data analysis. BayesB models were used to calibrate equations for each of the traits. Prediction accuracy was estimated for each trait and model using 20 replicates of a training-testing validation procedure. The repeatability of MCP measures and their predictions were also compared. The correlations between measured and predicted traits, in the external validation, were always higher than 0.5 (0.88 for rennet coagulation time). We confirmed that the most important element for finding the prediction accuracy is the repeatability of the gold standard analyses used for building calibration equations. Repeatability measures of the predicted traits were generally high (≥95%), even for those traits with moderate analytical repeatability. Our results show that Bayesian models applied to Fourier-transform infrared spectra are powerful tools for cheap and rapid prediction of important traits in ovine milk and, compared with other methods, could help in the interpretation of results. Copyright © 2017 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.

Bayesian decision support for coding occupational injury data.

PubMed

Nanda, Gaurav; Grattan, Kathleen M; Chu, MyDzung T; Davis, Letitia K; Lehto, Mark R

2016-06-01

Studies on autocoding injury data have found that machine learning algorithms perform well for categories that occur frequently but often struggle with rare categories. Therefore, manual coding, although resource-intensive, cannot be eliminated. We propose a Bayesian decision support system to autocode a large portion of the data, filter cases for manual review, and assist human coders by presenting them top k prediction choices and a confusion matrix of predictions from Bayesian models. We studied the prediction performance of Single-Word (SW) and Two-Word-Sequence (TW) Naïve Bayes models on a sample of data from the 2011 Survey of Occupational Injury and Illness (SOII). We used the agreement in prediction results of SW and TW models, and various prediction strength thresholds for autocoding and filtering cases for manual review. We also studied the sensitivity of the top k predictions of the SW model, TW model, and SW-TW combination, and then compared the accuracy of the manually assigned codes to SOII data with that of the proposed system. The accuracy of the proposed system, assuming well-trained coders reviewing a subset of only 26% of cases flagged for review, was estimated to be comparable (86.5%) to the accuracy of the original coding of the data set (range: 73%-86.8%). Overall, the TW model had higher sensitivity than the SW model, and the accuracy of the prediction results increased when the two models agreed, and for higher prediction strength thresholds. The sensitivity of the top five predictions was 93%. The proposed system seems promising for coding injury data as it offers comparable accuracy and less manual coding. Accurate and timely coded occupational injury data is useful for surveillance as well as prevention activities that aim to make workplaces safer. Copyright © 2016 Elsevier Ltd and National Safety Council. All rights reserved.

a Bayesian Synthesis of Predictions from Different Models for Setting Water Quality Criteria

NASA Astrophysics Data System (ADS)

Arhonditsis, G. B.; Ecological Modelling Laboratory

2011-12-01

Skeptical views of the scientific value of modelling argue that there is no true model of an ecological system, but rather several adequate descriptions of different conceptual basis and structure. In this regard, rather than picking the single "best-fit" model to predict future system responses, we can use Bayesian model averaging to synthesize the forecasts from different models. Hence, by acknowledging that models from different areas of the complexity spectrum have different strengths and weaknesses, the Bayesian model averaging is an appealing approach to improve the predictive capacity and to overcome the ambiguity surrounding the model selection or the risk of basing ecological forecasts on a single model. Our study addresses this question using a complex ecological model, developed by Ramin et al. (2011; Environ Modell Softw 26, 337-353) to guide the water quality criteria setting process in the Hamilton Harbour (Ontario, Canada), along with a simpler plankton model that considers the interplay among phosphate, detritus, and generic phytoplankton and zooplankton state variables. This simple approach is more easily subjected to detailed sensitivity analysis and also has the advantage of fewer unconstrained parameters. Using Markov Chain Monte Carlo simulations, we calculate the relative mean standard error to assess the posterior support of the two models from the existing data. Predictions from the two models are then combined using the respective standard error estimates as weights in a weighted model average. The model averaging approach is used to examine the robustness of predictive statements made from our earlier work regarding the response of Hamilton Harbour to the different nutrient loading reduction strategies. The two eutrophication models are then used in conjunction with the SPAtially Referenced Regressions On Watershed attributes (SPARROW) watershed model. The Bayesian nature of our work is used: (i) to alleviate problems of spatiotemporal resolution mismatch between watershed and receiving waterbody models; and (ii) to overcome the conceptual or scale misalignment between processes of interest and supporting information. The proposed Bayesian approach provides an effective means of empirically estimating the relation between in-stream measurements of nutrient fluxes and the sources/sinks of nutrients within the watershed, while explicitly accounting for the uncertainty associated with the existing knowledge from the system along with the different types of spatial correlation typically underlying the parameter estimation of watershed models. Our modelling exercise offers the first estimates of the export coefficients and the delivery rates from the different subcatchments and thus generates testable hypotheses regarding the nutrient export "hot spots" in the studied watershed. Finally, we conduct modeling experiments that evaluate the potential improvement of the model parameter estimates and the decrease of the predictive uncertainty, if the uncertainty associated with the contemporary nutrient loading estimates is reduced. The lessons learned from this study will contribute towards the development of integrated modelling frameworks.

Bayesian averaging over Decision Tree models for trauma severity scoring.

PubMed

Schetinin, V; Jakaite, L; Krzanowski, W

2018-01-01

Health care practitioners analyse possible risks of misleading decisions and need to estimate and quantify uncertainty in predictions. We have examined the "gold" standard of screening a patient's conditions for predicting survival probability, based on logistic regression modelling, which is used in trauma care for clinical purposes and quality audit. This methodology is based on theoretical assumptions about data and uncertainties. Models induced within such an approach have exposed a number of problems, providing unexplained fluctuation of predicted survival and low accuracy of estimating uncertainty intervals within which predictions are made. Bayesian method, which in theory is capable of providing accurate predictions and uncertainty estimates, has been adopted in our study using Decision Tree models. Our approach has been tested on a large set of patients registered in the US National Trauma Data Bank and has outperformed the standard method in terms of prediction accuracy, thereby providing practitioners with accurate estimates of the predictive posterior densities of interest that are required for making risk-aware decisions. Copyright © 2017 Elsevier B.V. All rights reserved.

Bayesian accounts of covert selective attention: A tutorial review.

PubMed

Vincent, Benjamin T

2015-05-01

Decision making and optimal observer models offer an important theoretical approach to the study of covert selective attention. While their probabilistic formulation allows quantitative comparison to human performance, the models can be complex and their insights are not always immediately apparent. Part 1 establishes the theoretical appeal of the Bayesian approach, and introduces the way in which probabilistic approaches can be applied to covert search paradigms. Part 2 presents novel formulations of Bayesian models of 4 important covert attention paradigms, illustrating optimal observer predictions over a range of experimental manipulations. Graphical model notation is used to present models in an accessible way and Supplementary Code is provided to help bridge the gap between model theory and practical implementation. Part 3 reviews a large body of empirical and modelling evidence showing that many experimental phenomena in the domain of covert selective attention are a set of by-products. These effects emerge as the result of observers conducting Bayesian inference with noisy sensory observations, prior expectations, and knowledge of the generative structure of the stimulus environment.

Bayesian Action–Perception Computational Model: Interaction of Production and Recognition of Cursive Letters

PubMed Central

Gilet, Estelle; Diard, Julien; Bessière, Pierre

2011-01-01

In this paper, we study the collaboration of perception and action representations involved in cursive letter recognition and production. We propose a mathematical formulation for the whole perception–action loop, based on probabilistic modeling and Bayesian inference, which we call the Bayesian Action–Perception (BAP) model. Being a model of both perception and action processes, the purpose of this model is to study the interaction of these processes. More precisely, the model includes a feedback loop from motor production, which implements an internal simulation of movement. Motor knowledge can therefore be involved during perception tasks. In this paper, we formally define the BAP model and show how it solves the following six varied cognitive tasks using Bayesian inference: i) letter recognition (purely sensory), ii) writer recognition, iii) letter production (with different effectors), iv) copying of trajectories, v) copying of letters, and vi) letter recognition (with internal simulation of movements). We present computer simulations of each of these cognitive tasks, and discuss experimental predictions and theoretical developments. PMID:21674043

Remaining useful life assessment of lithium-ion batteries in implantable medical devices

NASA Astrophysics Data System (ADS)

Hu, Chao; Ye, Hui; Jain, Gaurav; Schmidt, Craig

2018-01-01

This paper presents a prognostic study on lithium-ion batteries in implantable medical devices, in which a hybrid data-driven/model-based method is employed for remaining useful life assessment. The method is developed on and evaluated against data from two sets of lithium-ion prismatic cells used in implantable applications exhibiting distinct fade performance: 1) eight cells from Medtronic, PLC whose rates of capacity fade appear to be stable and gradually decrease over a 10-year test duration; and 2) eight cells from Manufacturer X whose rates appear to be greater and show sharp increase after some period over a 1.8-year test duration. The hybrid method enables online prediction of remaining useful life for predictive maintenance/control. It consists of two modules: 1) a sparse Bayesian learning module (data-driven) for inferring capacity from charge-related features; and 2) a recursive Bayesian filtering module (model-based) for updating empirical capacity fade models and predicting remaining useful life. A generic particle filter is adopted to implement recursive Bayesian filtering for the cells from the first set, whose capacity fade behavior can be represented by a single fade model; a multiple model particle filter with fixed-lag smoothing is proposed for the cells from the second data set, whose capacity fade behavior switches between multiple fade models.

Bayesian Parameter Inference and Model Selection by Population Annealing in Systems Biology

PubMed Central

Murakami, Yohei

2014-01-01

Parameter inference and model selection are very important for mathematical modeling in systems biology. Bayesian statistics can be used to conduct both parameter inference and model selection. Especially, the framework named approximate Bayesian computation is often used for parameter inference and model selection in systems biology. However, Monte Carlo methods needs to be used to compute Bayesian posterior distributions. In addition, the posterior distributions of parameters are sometimes almost uniform or very similar to their prior distributions. In such cases, it is difficult to choose one specific value of parameter with high credibility as the representative value of the distribution. To overcome the problems, we introduced one of the population Monte Carlo algorithms, population annealing. Although population annealing is usually used in statistical mechanics, we showed that population annealing can be used to compute Bayesian posterior distributions in the approximate Bayesian computation framework. To deal with un-identifiability of the representative values of parameters, we proposed to run the simulations with the parameter ensemble sampled from the posterior distribution, named “posterior parameter ensemble”. We showed that population annealing is an efficient and convenient algorithm to generate posterior parameter ensemble. We also showed that the simulations with the posterior parameter ensemble can, not only reproduce the data used for parameter inference, but also capture and predict the data which was not used for parameter inference. Lastly, we introduced the marginal likelihood in the approximate Bayesian computation framework for Bayesian model selection. We showed that population annealing enables us to compute the marginal likelihood in the approximate Bayesian computation framework and conduct model selection depending on the Bayes factor. PMID:25089832

Bayesian Scoring Systems for Military Pelvic and Perineal Blast Injuries: Is it Time to Take a New Approach?

PubMed

Mossadegh, Somayyeh; He, Shan; Parker, Paul

2016-05-01

Various injury severity scores exist for trauma; it is known that they do not correlate accurately to military injuries. A promising anatomical scoring system for blast pelvic and perineal injury led to the development of an improved scoring system using machine-learning techniques. An unbiased genetic algorithm selected optimal anatomical and physiological parameters from 118 military cases. A Naïve Bayesian model was built using the proposed parameters to predict the probability of survival. Ten-fold cross validation was employed to evaluate its performance. Our model significantly out-performed Injury Severity Score (ISS), Trauma ISS, New ISS, and the Revised Trauma Score in virtually all areas; positive predictive value 0.8941, specificity 0.9027, accuracy 0.9056, and area under curve 0.9059. A two-sample t test showed that the predictive performance of the proposed scoring system was significantly better than the other systems (p < 0.001). With limited resources and the simplest of Bayesian methodologies, we have demonstrated that the Naïve Bayesian model performed significantly better in virtually all areas assessed by current scoring systems used for trauma. This is encouraging and highlights that more can be done to improve trauma systems not only for our military injured, but also for civilian trauma victims. Reprint & Copyright © 2016 Association of Military Surgeons of the U.S.

With or without you: predictive coding and Bayesian inference in the brain

PubMed Central

Aitchison, Laurence; Lengyel, Máté

2018-01-01

Two theoretical ideas have emerged recently with the ambition to provide a unifying functional explanation of neural population coding and dynamics: predictive coding and Bayesian inference. Here, we describe the two theories and their combination into a single framework: Bayesian predictive coding. We clarify how the two theories can be distinguished, despite sharing core computational concepts and addressing an overlapping set of empirical phenomena. We argue that predictive coding is an algorithmic / representational motif that can serve several different computational goals of which Bayesian inference is but one. Conversely, while Bayesian inference can utilize predictive coding, it can also be realized by a variety of other representations. We critically evaluate the experimental evidence supporting Bayesian predictive coding and discuss how to test it more directly. PMID:28942084

A Bayesian network approach for modeling local failure in lung cancer

NASA Astrophysics Data System (ADS)

Oh, Jung Hun; Craft, Jeffrey; Lozi, Rawan Al; Vaidya, Manushka; Meng, Yifan; Deasy, Joseph O.; Bradley, Jeffrey D.; El Naqa, Issam

2011-03-01

Locally advanced non-small cell lung cancer (NSCLC) patients suffer from a high local failure rate following radiotherapy. Despite many efforts to develop new dose-volume models for early detection of tumor local failure, there was no reported significant improvement in their application prospectively. Based on recent studies of biomarker proteins' role in hypoxia and inflammation in predicting tumor response to radiotherapy, we hypothesize that combining physical and biological factors with a suitable framework could improve the overall prediction. To test this hypothesis, we propose a graphical Bayesian network framework for predicting local failure in lung cancer. The proposed approach was tested using two different datasets of locally advanced NSCLC patients treated with radiotherapy. The first dataset was collected retrospectively, which comprises clinical and dosimetric variables only. The second dataset was collected prospectively in which in addition to clinical and dosimetric information, blood was drawn from the patients at various time points to extract candidate biomarkers as well. Our preliminary results show that the proposed method can be used as an efficient method to develop predictive models of local failure in these patients and to interpret relationships among the different variables in the models. We also demonstrate the potential use of heterogeneous physical and biological variables to improve the model prediction. With the first dataset, we achieved better performance compared with competing Bayesian-based classifiers. With the second dataset, the combined model had a slightly higher performance compared to individual physical and biological models, with the biological variables making the largest contribution. Our preliminary results highlight the potential of the proposed integrated approach for predicting post-radiotherapy local failure in NSCLC patients.

Bayesian Action-Perception loop modeling: Application to trajectory generation and recognition using internal motor simulation

NASA Astrophysics Data System (ADS)

Gilet, Estelle; Diard, Julien; Palluel-Germain, Richard; Bessière, Pierre

2011-03-01

This paper is about modeling perception-action loops and, more precisely, the study of the influence of motor knowledge during perception tasks. We use the Bayesian Action-Perception (BAP) model, which deals with the sensorimotor loop involved in reading and writing cursive isolated letters and includes an internal simulation of movement loop. By using this probabilistic model we simulate letter recognition, both with and without internal motor simulation. Comparison of their performance yields an experimental prediction, which we set forth.

Optimization of Biomathematical Model Predictions for Cognitive Performance Impairment in Individuals: Accounting for Unknown Traits and Uncertain States in Homeostatic and Circadian Processes

PubMed Central

Van Dongen, Hans P. A.; Mott, Christopher G.; Huang, Jen-Kuang; Mollicone, Daniel J.; McKenzie, Frederic D.; Dinges, David F.

2007-01-01

Current biomathematical models of fatigue and performance do not accurately predict cognitive performance for individuals with a priori unknown degrees of trait vulnerability to sleep loss, do not predict performance reliably when initial conditions are uncertain, and do not yield statistically valid estimates of prediction accuracy. These limitations diminish their usefulness for predicting the performance of individuals in operational environments. To overcome these 3 limitations, a novel modeling approach was developed, based on the expansion of a statistical technique called Bayesian forecasting. The expanded Bayesian forecasting procedure was implemented in the two-process model of sleep regulation, which has been used to predict performance on the basis of the combination of a sleep homeostatic process and a circadian process. Employing the two-process model with the Bayesian forecasting procedure to predict performance for individual subjects in the face of unknown traits and uncertain states entailed subject-specific optimization of 3 trait parameters (homeostatic build-up rate, circadian amplitude, and basal performance level) and 2 initial state parameters (initial homeostatic state and circadian phase angle). Prior information about the distribution of the trait parameters in the population at large was extracted from psychomotor vigilance test (PVT) performance measurements in 10 subjects who had participated in a laboratory experiment with 88 h of total sleep deprivation. The PVT performance data of 3 additional subjects in this experiment were set aside beforehand for use in prospective computer simulations. The simulations involved updating the subject-specific model parameters every time the next performance measurement became available, and then predicting performance 24 h ahead. Comparison of the predictions to the subjects' actual data revealed that as more data became available for the individuals at hand, the performance predictions became increasingly more accurate and had progressively smaller 95% confidence intervals, as the model parameters converged efficiently to those that best characterized each individual. Even when more challenging simulations were run (mimicking a change in the initial homeostatic state; simulating the data to be sparse), the predictions were still considerably more accurate than would have been achieved by the two-process model alone. Although the work described here is still limited to periods of consolidated wakefulness with stable circadian rhythms, the results obtained thus far indicate that the Bayesian forecasting procedure can successfully overcome some of the major outstanding challenges for biomathematical prediction of cognitive performance in operational settings. Citation: Van Dongen HPA; Mott CG; Huang JK; Mollicone DJ; McKenzie FD; Dinges DF. Optimization of biomathematical model predictions for cognitive performance impairment in individuals: accounting for unknown traits and uncertain states in homeostatic and circadian processes. SLEEP 2007;30(9):1129-1143. PMID:17910385

Robust Bayesian Analysis of Heavy-tailed Stochastic Volatility Models using Scale Mixtures of Normal Distributions

PubMed Central

Abanto-Valle, C. A.; Bandyopadhyay, D.; Lachos, V. H.; Enriquez, I.

2009-01-01

A Bayesian analysis of stochastic volatility (SV) models using the class of symmetric scale mixtures of normal (SMN) distributions is considered. In the face of non-normality, this provides an appealing robust alternative to the routine use of the normal distribution. Specific distributions examined include the normal, student-t, slash and the variance gamma distributions. Using a Bayesian paradigm, an efficient Markov chain Monte Carlo (MCMC) algorithm is introduced for parameter estimation. Moreover, the mixing parameters obtained as a by-product of the scale mixture representation can be used to identify outliers. The methods developed are applied to analyze daily stock returns data on S&P500 index. Bayesian model selection criteria as well as out-of- sample forecasting results reveal that the SV models based on heavy-tailed SMN distributions provide significant improvement in model fit as well as prediction to the S&P500 index data over the usual normal model. PMID:20730043

Bayesian neural adjustment of inhibitory control predicts emergence of problem stimulant use.

PubMed

Harlé, Katia M; Stewart, Jennifer L; Zhang, Shunan; Tapert, Susan F; Yu, Angela J; Paulus, Martin P

2015-11-01

Bayesian ideal observer models quantify individuals' context- and experience-dependent beliefs and expectations about their environment, which provides a powerful approach (i) to link basic behavioural mechanisms to neural processing; and (ii) to generate clinical predictors for patient populations. Here, we focus on (ii) and determine whether individual differences in the neural representation of the need to stop in an inhibitory task can predict the development of problem use (i.e. abuse or dependence) in individuals experimenting with stimulants. One hundred and fifty-seven non-dependent occasional stimulant users, aged 18-24, completed a stop-signal task while undergoing functional magnetic resonance imaging. These individuals were prospectively followed for 3 years and evaluated for stimulant use and abuse/dependence symptoms. At follow-up, 38 occasional stimulant users met criteria for a stimulant use disorder (problem stimulant users), while 50 had discontinued use (desisted stimulant users). We found that those individuals who showed greater neural responses associated with Bayesian prediction errors, i.e. the difference between actual and expected need to stop on a given trial, in right medial prefrontal cortex/anterior cingulate cortex, caudate, anterior insula, and thalamus were more likely to exhibit problem use 3 years later. Importantly, these computationally based neural predictors outperformed clinical measures and non-model based neural variables in predicting clinical status. In conclusion, young adults who show exaggerated brain processing underlying whether to 'stop' or to 'go' are more likely to develop stimulant abuse. Thus, Bayesian cognitive models provide both a computational explanation and potential predictive biomarkers of belief processing deficits in individuals at risk for stimulant addiction. © The Author (2015). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

A Bayesian alternative for multi-objective ecohydrological model specification

NASA Astrophysics Data System (ADS)

Tang, Yating; Marshall, Lucy; Sharma, Ashish; Ajami, Hoori

2018-01-01

Recent studies have identified the importance of vegetation processes in terrestrial hydrologic systems. Process-based ecohydrological models combine hydrological, physical, biochemical and ecological processes of the catchments, and as such are generally more complex and parametric than conceptual hydrological models. Thus, appropriate calibration objectives and model uncertainty analysis are essential for ecohydrological modeling. In recent years, Bayesian inference has become one of the most popular tools for quantifying the uncertainties in hydrological modeling with the development of Markov chain Monte Carlo (MCMC) techniques. The Bayesian approach offers an appealing alternative to traditional multi-objective hydrologic model calibrations by defining proper prior distributions that can be considered analogous to the ad-hoc weighting often prescribed in multi-objective calibration. Our study aims to develop appropriate prior distributions and likelihood functions that minimize the model uncertainties and bias within a Bayesian ecohydrological modeling framework based on a traditional Pareto-based model calibration technique. In our study, a Pareto-based multi-objective optimization and a formal Bayesian framework are implemented in a conceptual ecohydrological model that combines a hydrological model (HYMOD) and a modified Bucket Grassland Model (BGM). Simulations focused on one objective (streamflow/LAI) and multiple objectives (streamflow and LAI) with different emphasis defined via the prior distribution of the model error parameters. Results show more reliable outputs for both predicted streamflow and LAI using Bayesian multi-objective calibration with specified prior distributions for error parameters based on results from the Pareto front in the ecohydrological modeling. The methodology implemented here provides insight into the usefulness of multiobjective Bayesian calibration for ecohydrologic systems and the importance of appropriate prior distributions in such approaches.

A Bayesian approach to estimate the biomass of anchovies off the coast of Perú.

PubMed

Quiroz, Zaida C; Prates, Marcos O; Rue, Håvard

2015-03-01

The Northern Humboldt Current System (NHCS) is the world's most productive ecosystem in terms of fish. In particular, the Peruvian anchovy (Engraulis ringens) is the major prey of the main top predators, like seabirds, fish, humans, and other mammals. In this context, it is important to understand the dynamics of the anchovy distribution to preserve it as well as to exploit its economic capacities. Using the data collected by the "Instituto del Mar del Perú" (IMARPE) during a scientific survey in 2005, we present a statistical analysis that has as main goals: (i) to adapt to the characteristics of the sampled data, such as spatial dependence, high proportions of zeros and big size of samples; (ii) to provide important insights on the dynamics of the anchovy population; and (iii) to propose a model for estimation and prediction of anchovy biomass in the NHCS offshore from Perú. These data were analyzed in a Bayesian framework using the integrated nested Laplace approximation (INLA) method. Further, to select the best model and to study the predictive power of each model, we performed model comparisons and predictive checks, respectively. Finally, we carried out a Bayesian spatial influence diagnostic for the preferred model. © 2014, The International Biometric Society.

Prediction of road accidents: A Bayesian hierarchical approach.

PubMed

Deublein, Markus; Schubert, Matthias; Adey, Bryan T; Köhler, Jochen; Faber, Michael H

2013-03-01

In this paper a novel methodology for the prediction of the occurrence of road accidents is presented. The methodology utilizes a combination of three statistical methods: (1) gamma-updating of the occurrence rates of injury accidents and injured road users, (2) hierarchical multivariate Poisson-lognormal regression analysis taking into account correlations amongst multiple dependent model response variables and effects of discrete accident count data e.g. over-dispersion, and (3) Bayesian inference algorithms, which are applied by means of data mining techniques supported by Bayesian Probabilistic Networks in order to represent non-linearity between risk indicating and model response variables, as well as different types of uncertainties which might be present in the development of the specific models. Prior Bayesian Probabilistic Networks are first established by means of multivariate regression analysis of the observed frequencies of the model response variables, e.g. the occurrence of an accident, and observed values of the risk indicating variables, e.g. degree of road curvature. Subsequently, parameter learning is done using updating algorithms, to determine the posterior predictive probability distributions of the model response variables, conditional on the values of the risk indicating variables. The methodology is illustrated through a case study using data of the Austrian rural motorway network. In the case study, on randomly selected road segments the methodology is used to produce a model to predict the expected number of accidents in which an injury has occurred and the expected number of light, severe and fatally injured road users. Additionally, the methodology is used for geo-referenced identification of road sections with increased occurrence probabilities of injury accident events on a road link between two Austrian cities. It is shown that the proposed methodology can be used to develop models to estimate the occurrence of road accidents for any road network provided that the required data are available. Copyright © 2012 Elsevier Ltd. All rights reserved.

Predicting forest insect flight activity: A Bayesian network approach

Treesearch

Stephen M. Pawson; Bruce G. Marcot; Owen G. Woodberry

2017-01-01

Daily flight activity patterns of forest insects are influenced by temporal and meteorological conditions. Temperature and time of day are frequently cited as key drivers of activity; however, complex interactions between multiple contributing factors have also been proposed. Here, we report individual Bayesian network models to assess the probability of flight...

A Comparison of Imputation Methods for Bayesian Factor Analysis Models

ERIC Educational Resources Information Center

Merkle, Edgar C.

2011-01-01

Imputation methods are popular for the handling of missing data in psychology. The methods generally consist of predicting missing data based on observed data, yielding a complete data set that is amiable to standard statistical analyses. In the context of Bayesian factor analysis, this article compares imputation under an unrestricted…

Optimal population prediction of sandhill crane recruitment based on climate-mediated habitat limitations

USGS Publications Warehouse

Gerber, Brian D.; Kendall, William L.; Hooten, Mevin B.; Dubovsky, James A.; Drewien, Roderick C.

2015-01-01

Prediction is fundamental to scientific enquiry and application; however, ecologists tend to favour explanatory modelling. We discuss a predictive modelling framework to evaluate ecological hypotheses and to explore novel/unobserved environmental scenarios to assist conservation and management decision-makers. We apply this framework to develop an optimal predictive model for juvenile (<1 year old) sandhill crane Grus canadensis recruitment of the Rocky Mountain Population (RMP). We consider spatial climate predictors motivated by hypotheses of how drought across multiple time-scales and spring/summer weather affects recruitment.Our predictive modelling framework focuses on developing a single model that includes all relevant predictor variables, regardless of collinearity. This model is then optimized for prediction by controlling model complexity using a data-driven approach that marginalizes or removes irrelevant predictors from the model. Specifically, we highlight two approaches of statistical regularization, Bayesian least absolute shrinkage and selection operator (LASSO) and ridge regression.Our optimal predictive Bayesian LASSO and ridge regression models were similar and on average 37% superior in predictive accuracy to an explanatory modelling approach. Our predictive models confirmed a priori hypotheses that drought and cold summers negatively affect juvenile recruitment in the RMP. The effects of long-term drought can be alleviated by short-term wet spring–summer months; however, the alleviation of long-term drought has a much greater positive effect on juvenile recruitment. The number of freezing days and snowpack during the summer months can also negatively affect recruitment, while spring snowpack has a positive effect.Breeding habitat, mediated through climate, is a limiting factor on population growth of sandhill cranes in the RMP, which could become more limiting with a changing climate (i.e. increased drought). These effects are likely not unique to cranes. The alteration of hydrological patterns and water levels by drought may impact many migratory, wetland nesting birds in the Rocky Mountains and beyond.Generalizable predictive models (trained by out-of-sample fit and based on ecological hypotheses) are needed by conservation and management decision-makers. Statistical regularization improves predictions and provides a general framework for fitting models with a large number of predictors, even those with collinearity, to simultaneously identify an optimal predictive model while conducting rigorous Bayesian model selection. Our framework is important for understanding population dynamics under a changing climate and has direct applications for making harvest and habitat management decisions.

Bayesian analysis of non-linear differential equation models with application to a gut microbial ecosystem.

PubMed

Lawson, Daniel J; Holtrop, Grietje; Flint, Harry

2011-07-01

Process models specified by non-linear dynamic differential equations contain many parameters, which often must be inferred from a limited amount of data. We discuss a hierarchical Bayesian approach combining data from multiple related experiments in a meaningful way, which permits more powerful inference than treating each experiment as independent. The approach is illustrated with a simulation study and example data from experiments replicating the aspects of the human gut microbial ecosystem. A predictive model is obtained that contains prediction uncertainty caused by uncertainty in the parameters, and we extend the model to capture situations of interest that cannot easily be studied experimentally. Copyright © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

An Application of Bayesian Approach in Modeling Risk of Death in an Intensive Care Unit

PubMed Central

Wong, Rowena Syn Yin; Ismail, Noor Azina

2016-01-01

Background and Objectives There are not many studies that attempt to model intensive care unit (ICU) risk of death in developing countries, especially in South East Asia. The aim of this study was to propose and describe application of a Bayesian approach in modeling in-ICU deaths in a Malaysian ICU. Methods This was a prospective study in a mixed medical-surgery ICU in a multidisciplinary tertiary referral hospital in Malaysia. Data collection included variables that were defined in Acute Physiology and Chronic Health Evaluation IV (APACHE IV) model. Bayesian Markov Chain Monte Carlo (MCMC) simulation approach was applied in the development of four multivariate logistic regression predictive models for the ICU, where the main outcome measure was in-ICU mortality risk. The performance of the models were assessed through overall model fit, discrimination and calibration measures. Results from the Bayesian models were also compared against results obtained using frequentist maximum likelihood method. Results The study involved 1,286 consecutive ICU admissions between January 1, 2009 and June 30, 2010, of which 1,111 met the inclusion criteria. Patients who were admitted to the ICU were generally younger, predominantly male, with low co-morbidity load and mostly under mechanical ventilation. The overall in-ICU mortality rate was 18.5% and the overall mean Acute Physiology Score (APS) was 68.5. All four models exhibited good discrimination, with area under receiver operating characteristic curve (AUC) values approximately 0.8. Calibration was acceptable (Hosmer-Lemeshow p-values > 0.05) for all models, except for model M3. Model M1 was identified as the model with the best overall performance in this study. Conclusion Four prediction models were proposed, where the best model was chosen based on its overall performance in this study. This study has also demonstrated the promising potential of the Bayesian MCMC approach as an alternative in the analysis and modeling of in-ICU mortality outcomes. PMID:27007413

An Application of Bayesian Approach in Modeling Risk of Death in an Intensive Care Unit.

PubMed

Wong, Rowena Syn Yin; Ismail, Noor Azina

2016-01-01

There are not many studies that attempt to model intensive care unit (ICU) risk of death in developing countries, especially in South East Asia. The aim of this study was to propose and describe application of a Bayesian approach in modeling in-ICU deaths in a Malaysian ICU. This was a prospective study in a mixed medical-surgery ICU in a multidisciplinary tertiary referral hospital in Malaysia. Data collection included variables that were defined in Acute Physiology and Chronic Health Evaluation IV (APACHE IV) model. Bayesian Markov Chain Monte Carlo (MCMC) simulation approach was applied in the development of four multivariate logistic regression predictive models for the ICU, where the main outcome measure was in-ICU mortality risk. The performance of the models were assessed through overall model fit, discrimination and calibration measures. Results from the Bayesian models were also compared against results obtained using frequentist maximum likelihood method. The study involved 1,286 consecutive ICU admissions between January 1, 2009 and June 30, 2010, of which 1,111 met the inclusion criteria. Patients who were admitted to the ICU were generally younger, predominantly male, with low co-morbidity load and mostly under mechanical ventilation. The overall in-ICU mortality rate was 18.5% and the overall mean Acute Physiology Score (APS) was 68.5. All four models exhibited good discrimination, with area under receiver operating characteristic curve (AUC) values approximately 0.8. Calibration was acceptable (Hosmer-Lemeshow p-values > 0.05) for all models, except for model M3. Model M1 was identified as the model with the best overall performance in this study. Four prediction models were proposed, where the best model was chosen based on its overall performance in this study. This study has also demonstrated the promising potential of the Bayesian MCMC approach as an alternative in the analysis and modeling of in-ICU mortality outcomes.

Comparing models for quantitative risk assessment: an application to the European Registry of foreign body injuries in children.

PubMed

Berchialla, Paola; Scarinzi, Cecilia; Snidero, Silvia; Gregori, Dario

2016-08-01

Risk Assessment is the systematic study of decisions subject to uncertain consequences. An increasing interest has been focused on modeling techniques like Bayesian Networks since their capability of (1) combining in the probabilistic framework different type of evidence including both expert judgments and objective data; (2) overturning previous beliefs in the light of the new information being received and (3) making predictions even with incomplete data. In this work, we proposed a comparison among Bayesian Networks and other classical Quantitative Risk Assessment techniques such as Neural Networks, Classification Trees, Random Forests and Logistic Regression models. Hybrid approaches, combining both Classification Trees and Bayesian Networks, were also considered. Among Bayesian Networks, a clear distinction between purely data-driven approach and combination of expert knowledge with objective data is made. The aim of this paper consists in evaluating among this models which best can be applied, in the framework of Quantitative Risk Assessment, to assess the safety of children who are exposed to the risk of inhalation/insertion/aspiration of consumer products. The issue of preventing injuries in children is of paramount importance, in particular where product design is involved: quantifying the risk associated to product characteristics can be of great usefulness in addressing the product safety design regulation. Data of the European Registry of Foreign Bodies Injuries formed the starting evidence for risk assessment. Results showed that Bayesian Networks appeared to have both the ease of interpretability and accuracy in making prediction, even if simpler models like logistic regression still performed well. © The Author(s) 2013.

The Neural Correlates of Hierarchical Predictions for Perceptual Decisions.

PubMed

Weilnhammer, Veith A; Stuke, Heiner; Sterzer, Philipp; Schmack, Katharina

2018-05-23

Sensory information is inherently noisy, sparse, and ambiguous. In contrast, visual experience is usually clear, detailed, and stable. Bayesian theories of perception resolve this discrepancy by assuming that prior knowledge about the causes underlying sensory stimulation actively shapes perceptual decisions. The CNS is believed to entertain a generative model aligned to dynamic changes in the hierarchical states of our volatile sensory environment. Here, we used model-based fMRI to study the neural correlates of the dynamic updating of hierarchically structured predictions in male and female human observers. We devised a crossmodal associative learning task with covertly interspersed ambiguous trials in which participants engaged in hierarchical learning based on changing contingencies between auditory cues and visual targets. By inverting a Bayesian model of perceptual inference, we estimated individual hierarchical predictions, which significantly biased perceptual decisions under ambiguity. Although "high-level" predictions about the cue-target contingency correlated with activity in supramodal regions such as orbitofrontal cortex and hippocampus, dynamic "low-level" predictions about the conditional target probabilities were associated with activity in retinotopic visual cortex. Our results suggest that our CNS updates distinct representations of hierarchical predictions that continuously affect perceptual decisions in a dynamically changing environment. SIGNIFICANCE STATEMENT Bayesian theories posit that our brain entertains a generative model to provide hierarchical predictions regarding the causes of sensory information. Here, we use behavioral modeling and fMRI to study the neural underpinnings of such hierarchical predictions. We show that "high-level" predictions about the strength of dynamic cue-target contingencies during crossmodal associative learning correlate with activity in orbitofrontal cortex and the hippocampus, whereas "low-level" conditional target probabilities were reflected in retinotopic visual cortex. Our findings empirically corroborate theorizations on the role of hierarchical predictions in visual perception and contribute substantially to a longstanding debate on the link between sensory predictions and orbitofrontal or hippocampal activity. Our work fundamentally advances the mechanistic understanding of perceptual inference in the human brain. Copyright © 2018 the authors 0270-6474/18/385008-14$15.00/0.

Bayesian logistic regression approaches to predict incorrect DRG assignment.

PubMed

Suleiman, Mani; Demirhan, Haydar; Boyd, Leanne; Girosi, Federico; Aksakalli, Vural

2018-05-07

Episodes of care involving similar diagnoses and treatments and requiring similar levels of resource utilisation are grouped to the same Diagnosis-Related Group (DRG). In jurisdictions which implement DRG based payment systems, DRGs are a major determinant of funding for inpatient care. Hence, service providers often dedicate auditing staff to the task of checking that episodes have been coded to the correct DRG. The use of statistical models to estimate an episode's probability of DRG error can significantly improve the efficiency of clinical coding audits. This study implements Bayesian logistic regression models with weakly informative prior distributions to estimate the likelihood that episodes require a DRG revision, comparing these models with each other and to classical maximum likelihood estimates. All Bayesian approaches had more stable model parameters than maximum likelihood. The best performing Bayesian model improved overall classification per- formance by 6% compared to maximum likelihood, with a 34% gain compared to random classification, respectively. We found that the original DRG, coder and the day of coding all have a significant effect on the likelihood of DRG error. Use of Bayesian approaches has improved model parameter stability and classification accuracy. This method has already lead to improved audit efficiency in an operational capacity.

Bayesian Knowledge Fusion in Prognostics and Health Management—A Case Study

NASA Astrophysics Data System (ADS)

Rabiei, Masoud; Modarres, Mohammad; Mohammad-Djafari, Ali

2011-03-01

In the past few years, a research effort has been in progress at University of Maryland to develop a Bayesian framework based on Physics of Failure (PoF) for risk assessment and fleet management of aging airframes. Despite significant achievements in modelling of crack growth behavior using fracture mechanics, it is still of great interest to find practical techniques for monitoring the crack growth instances using nondestructive inspection and to integrate such inspection results with the fracture mechanics models to improve the predictions. The ultimate goal of this effort is to develop an integrated probabilistic framework for utilizing all of the available information to come up with enhanced (less uncertain) predictions for structural health of the aircraft in future missions. Such information includes material level fatigue models and test data, health monitoring measurements and inspection field data. In this paper, a case study of using Bayesian fusion technique for integrating information from multiple sources in a structural health management problem is presented.

Constructive Epistemic Modeling: A Hierarchical Bayesian Model Averaging Method

NASA Astrophysics Data System (ADS)

Tsai, F. T. C.; Elshall, A. S.

2014-12-01

Constructive epistemic modeling is the idea that our understanding of a natural system through a scientific model is a mental construct that continually develops through learning about and from the model. Using the hierarchical Bayesian model averaging (HBMA) method [1], this study shows that segregating different uncertain model components through a BMA tree of posterior model probabilities, model prediction, within-model variance, between-model variance and total model variance serves as a learning tool [2]. First, the BMA tree of posterior model probabilities permits the comparative evaluation of the candidate propositions of each uncertain model component. Second, systemic model dissection is imperative for understanding the individual contribution of each uncertain model component to the model prediction and variance. Third, the hierarchical representation of the between-model variance facilitates the prioritization of the contribution of each uncertain model component to the overall model uncertainty. We illustrate these concepts using the groundwater modeling of a siliciclastic aquifer-fault system. The sources of uncertainty considered are from geological architecture, formation dip, boundary conditions and model parameters. The study shows that the HBMA analysis helps in advancing knowledge about the model rather than forcing the model to fit a particularly understanding or merely averaging several candidate models. [1] Tsai, F. T.-C., and A. S. Elshall (2013), Hierarchical Bayesian model averaging for hydrostratigraphic modeling: Uncertainty segregation and comparative evaluation. Water Resources Research, 49, 5520-5536, doi:10.1002/wrcr.20428. [2] Elshall, A.S., and F. T.-C. Tsai (2014). Constructive epistemic modeling of groundwater flow with geological architecture and boundary condition uncertainty under Bayesian paradigm, Journal of Hydrology, 517, 105-119, doi: 10.1016/j.jhydrol.2014.05.027.

Accounting for Slipping and Other False Negatives in Logistic Models of Student Learning

ERIC Educational Resources Information Center

MacLellan, Christopher J.; Liu, Ran; Koedinger, Kenneth R.

2015-01-01

Additive Factors Model (AFM) and Performance Factors Analysis (PFA) are two popular models of student learning that employ logistic regression to estimate parameters and predict performance. This is in contrast to Bayesian Knowledge Tracing (BKT) which uses a Hidden Markov Model formalism. While all three models tend to make similar predictions,…

Advances in Applications of Hierarchical Bayesian Methods with Hydrological Models

NASA Astrophysics Data System (ADS)

Alexander, R. B.; Schwarz, G. E.; Boyer, E. W.

2017-12-01

Mechanistic and empirical watershed models are increasingly used to inform water resource decisions. Growing access to historical stream measurements and data from in-situ sensor technologies has increased the need for improved techniques for coupling models with hydrological measurements. Techniques that account for the intrinsic uncertainties of both models and measurements are especially needed. Hierarchical Bayesian methods provide an efficient modeling tool for quantifying model and prediction uncertainties, including those associated with measurements. Hierarchical methods can also be used to explore spatial and temporal variations in model parameters and uncertainties that are informed by hydrological measurements. We used hierarchical Bayesian methods to develop a hybrid (statistical-mechanistic) SPARROW (SPAtially Referenced Regression On Watershed attributes) model of long-term mean annual streamflow across diverse environmental and climatic drainages in 18 U.S. hydrological regions. Our application illustrates the use of a new generation of Bayesian methods that offer more advanced computational efficiencies than the prior generation. Evaluations of the effects of hierarchical (regional) variations in model coefficients and uncertainties on model accuracy indicates improved prediction accuracies (median of 10-50%) but primarily in humid eastern regions, where model uncertainties are one-third of those in arid western regions. Generally moderate regional variability is observed for most hierarchical coefficients. Accounting for measurement and structural uncertainties, using hierarchical state-space techniques, revealed the effects of spatially-heterogeneous, latent hydrological processes in the "localized" drainages between calibration sites; this improved model precision, with only minor changes in regional coefficients. Our study can inform advances in the use of hierarchical methods with hydrological models to improve their integration with stream measurements.

Bayesian hierarchical model of ceftriaxone resistance proportions among Salmonella serotype Heidelberg infections.

PubMed

Gu, Weidong; Medalla, Felicita; Hoekstra, Robert M

2018-02-01

The National Antimicrobial Resistance Monitoring System (NARMS) at the Centers for Disease Control and Prevention tracks resistance among Salmonella infections. The annual number of Salmonella isolates of a particular serotype from states may be small, making direct estimation of resistance proportions unreliable. We developed a Bayesian hierarchical model to improve estimation by borrowing strength from relevant sampling units. We illustrate the models with different specifications of spatio-temporal interaction using 2004-2013 NARMS data for ceftriaxone-resistant Salmonella serotype Heidelberg. Our results show that Bayesian estimates of resistance proportions were smoother than observed values, and the difference between predicted and observed proportions was inversely related to the number of submitted isolates. The model with interaction allowed for tracking of annual changes in resistance proportions at the state level. We demonstrated that Bayesian hierarchical models provide a useful tool to examine spatio-temporal patterns of small sample size such as those found in NARMS. Published by Elsevier Ltd.

A combined Fuzzy and Naive Bayesian strategy can be used to assign event codes to injury narratives.

PubMed

Marucci-Wellman, H; Lehto, M; Corns, H

2011-12-01

Bayesian methods show promise for classifying injury narratives from large administrative datasets into cause groups. This study examined a combined approach where two Bayesian models (Fuzzy and Naïve) were used to either classify a narrative or select it for manual review. Injury narratives were extracted from claims filed with a worker's compensation insurance provider between January 2002 and December 2004. Narratives were separated into a training set (n=11,000) and prediction set (n=3,000). Expert coders assigned two-digit Bureau of Labor Statistics Occupational Injury and Illness Classification event codes to each narrative. Fuzzy and Naïve Bayesian models were developed using manually classified cases in the training set. Two semi-automatic machine coding strategies were evaluated. The first strategy assigned cases for manual review if the Fuzzy and Naïve models disagreed on the classification. The second strategy selected additional cases for manual review from the Agree dataset using prediction strength to reach a level of 50% computer coding and 50% manual coding. When agreement alone was used as the filtering strategy, the majority were coded by the computer (n=1,928, 64%) leaving 36% for manual review. The overall combined (human plus computer) sensitivity was 0.90 and positive predictive value (PPV) was >0.90 for 11 of 18 2-digit event categories. Implementing the 2nd strategy improved results with an overall sensitivity of 0.95 and PPV >0.90 for 17 of 18 categories. A combined Naïve-Fuzzy Bayesian approach can classify some narratives with high accuracy and identify others most beneficial for manual review, reducing the burden on human coders.

Integrating GLL-Weibull Distribution Within a Bayesian Framework for Life Prediction of Shape Memory Alloy Spring Undergoing Thermo-mechanical Fatigue

NASA Astrophysics Data System (ADS)

Kundu, Pradeep; Nath, Tameshwer; Palani, I. A.; Lad, Bhupesh K.

2018-06-01

The present paper tackles an important but unmapped problem of the reliability estimations of smart materials. First, an experimental setup is developed for accelerated life testing of the shape memory alloy (SMA) springs. Generalized log-linear Weibull (GLL-Weibull) distribution-based novel approach is then developed for SMA spring life estimation. Applied stimulus (voltage), elongation and cycles of operation are used as inputs for the life prediction model. The values of the parameter coefficients of the model provide better interpretability compared to artificial intelligence based life prediction approaches. In addition, the model also considers the effect of operating conditions, making it generic for a range of the operating conditions. Moreover, a Bayesian framework is used to continuously update the prediction with the actual degradation value of the springs, thereby reducing the uncertainty in the data and improving the prediction accuracy. In addition, the deterioration of material with number of cycles is also investigated using thermogravimetric analysis and scanning electron microscopy.

Population pharmacokinetics and maximum a posteriori probability Bayesian estimator of abacavir: application of individualized therapy in HIV-infected infants and toddlers

PubMed Central

Zhao, Wei; Cella, Massimo; Della Pasqua, Oscar; Burger, David; Jacqz-Aigrain, Evelyne

2012-01-01

AIMS To develop a population pharmacokinetic model for abacavir in HIV-infected infants and toddlers, which will be used to describe both once and twice daily pharmacokinetic profiles, identify covariates that explain variability and propose optimal time points to optimize the area under the concentration–time curve (AUC) targeted dosage and individualize therapy. METHODS The pharmacokinetics of abacavir was described with plasma concentrations from 23 patients using nonlinear mixed-effects modelling (NONMEM) software. A two-compartment model with first-order absorption and elimination was developed. The final model was validated using bootstrap, visual predictive check and normalized prediction distribution errors. The Bayesian estimator was validated using the cross-validation and simulation–estimation method. RESULTS The typical population pharmacokinetic parameters and relative standard errors (RSE) were apparent systemic clearance (CL) 13.4 l h−1 (RSE 6.3%), apparent central volume of distribution 4.94 l (RSE 28.7%), apparent peripheral volume of distribution 8.12 l (RSE14.2%), apparent intercompartment clearance 1.25 l h−1 (RSE 16.9%) and absorption rate constant 0.758 h−1 (RSE 5.8%). The covariate analysis identified weight as the individual factor influencing the apparent oral clearance: CL = 13.4 × (weight/12)1.14. The maximum a posteriori probability Bayesian estimator, based on three concentrations measured at 0, 1 or 2, and 3 h after drug intake allowed predicting individual AUC0–t. CONCLUSIONS The population pharmacokinetic model developed for abacavir in HIV-infected infants and toddlers accurately described both once and twice daily pharmacokinetic profiles. The maximum a posteriori probability Bayesian estimator of AUC0–t was developed from the final model and can be used routinely to optimize individual dosing. PMID:21988586

Evaluating scaling models in biology using hierarchical Bayesian approaches

PubMed Central

Price, Charles A; Ogle, Kiona; White, Ethan P; Weitz, Joshua S

2009-01-01

Theoretical models for allometric relationships between organismal form and function are typically tested by comparing a single predicted relationship with empirical data. Several prominent models, however, predict more than one allometric relationship, and comparisons among alternative models have not taken this into account. Here we evaluate several different scaling models of plant morphology within a hierarchical Bayesian framework that simultaneously fits multiple scaling relationships to three large allometric datasets. The scaling models include: inflexible universal models derived from biophysical assumptions (e.g. elastic similarity or fractal networks), a flexible variation of a fractal network model, and a highly flexible model constrained only by basic algebraic relationships. We demonstrate that variation in intraspecific allometric scaling exponents is inconsistent with the universal models, and that more flexible approaches that allow for biological variability at the species level outperform universal models, even when accounting for relative increases in model complexity. PMID:19453621

Toward improved prediction of the bedrock depth underneath hillslopes: Bayesian inference of the bottom-up control hypothesis using high-resolution topographic data

NASA Astrophysics Data System (ADS)

Gomes, Guilherme J. C.; Vrugt, Jasper A.; Vargas, Eurípedes A.

2016-04-01

The depth to bedrock controls a myriad of processes by influencing subsurface flow paths, erosion rates, soil moisture, and water uptake by plant roots. As hillslope interiors are very difficult and costly to illuminate and access, the topography of the bedrock surface is largely unknown. This essay is concerned with the prediction of spatial patterns in the depth to bedrock (DTB) using high-resolution topographic data, numerical modeling, and Bayesian analysis. Our DTB model builds on the bottom-up control on fresh-bedrock topography hypothesis of Rempe and Dietrich (2014) and includes a mass movement and bedrock-valley morphology term to extent the usefulness and general applicability of the model. We reconcile the DTB model with field observations using Bayesian analysis with the DREAM algorithm. We investigate explicitly the benefits of using spatially distributed parameter values to account implicitly, and in a relatively simple way, for rock mass heterogeneities that are very difficult, if not impossible, to characterize adequately in the field. We illustrate our method using an artificial data set of bedrock depth observations and then evaluate our DTB model with real-world data collected at the Papagaio river basin in Rio de Janeiro, Brazil. Our results demonstrate that the DTB model predicts accurately the observed bedrock depth data. The posterior mean DTB simulation is shown to be in good agreement with the measured data. The posterior prediction uncertainty of the DTB model can be propagated forward through hydromechanical models to derive probabilistic estimates of factors of safety.

Bayesian energy landscape tilting: towards concordant models of molecular ensembles.

PubMed

Beauchamp, Kyle A; Pande, Vijay S; Das, Rhiju

2014-03-18

Predicting biological structure has remained challenging for systems such as disordered proteins that take on myriad conformations. Hybrid simulation/experiment strategies have been undermined by difficulties in evaluating errors from computational model inaccuracies and data uncertainties. Building on recent proposals from maximum entropy theory and nonequilibrium thermodynamics, we address these issues through a Bayesian energy landscape tilting (BELT) scheme for computing Bayesian hyperensembles over conformational ensembles. BELT uses Markov chain Monte Carlo to directly sample maximum-entropy conformational ensembles consistent with a set of input experimental observables. To test this framework, we apply BELT to model trialanine, starting from disagreeing simulations with the force fields ff96, ff99, ff99sbnmr-ildn, CHARMM27, and OPLS-AA. BELT incorporation of limited chemical shift and (3)J measurements gives convergent values of the peptide's α, β, and PPII conformational populations in all cases. As a test of predictive power, all five BELT hyperensembles recover set-aside measurements not used in the fitting and report accurate errors, even when starting from highly inaccurate simulations. BELT's principled framework thus enables practical predictions for complex biomolecular systems from discordant simulations and sparse data. Copyright © 2014 Biophysical Society. Published by Elsevier Inc. All rights reserved.

Predicting forest insect flight activity: A Bayesian network approach

PubMed Central

Pawson, Stephen M.; Marcot, Bruce G.; Woodberry, Owen G.

2017-01-01

Daily flight activity patterns of forest insects are influenced by temporal and meteorological conditions. Temperature and time of day are frequently cited as key drivers of activity; however, complex interactions between multiple contributing factors have also been proposed. Here, we report individual Bayesian network models to assess the probability of flight activity of three exotic insects, Hylurgus ligniperda, Hylastes ater, and Arhopalus ferus in a managed plantation forest context. Models were built from 7,144 individual hours of insect sampling, temperature, wind speed, relative humidity, photon flux density, and temporal data. Discretized meteorological and temporal variables were used to build naïve Bayes tree augmented networks. Calibration results suggested that the H. ater and A. ferus Bayesian network models had the best fit for low Type I and overall errors, and H. ligniperda had the best fit for low Type II errors. Maximum hourly temperature and time since sunrise had the largest influence on H. ligniperda flight activity predictions, whereas time of day and year had the greatest influence on H. ater and A. ferus activity. Type II model errors for the prediction of no flight activity is improved by increasing the model’s predictive threshold. Improvements in model performance can be made by further sampling, increasing the sensitivity of the flight intercept traps, and replicating sampling in other regions. Predicting insect flight informs an assessment of the potential phytosanitary risks of wood exports. Quantifying this risk allows mitigation treatments to be targeted to prevent the spread of invasive species via international trade pathways. PMID:28953904

Combining cow and bull reference populations to increase accuracy of genomic prediction and genome-wide association studies.

PubMed

Calus, M P L; de Haas, Y; Veerkamp, R F

2013-10-01

Genomic selection holds the promise to be particularly beneficial for traits that are difficult or expensive to measure, such that access to phenotypes on large daughter groups of bulls is limited. Instead, cow reference populations can be generated, potentially supplemented with existing information from the same or (highly) correlated traits available on bull reference populations. The objective of this study, therefore, was to develop a model to perform genomic predictions and genome-wide association studies based on a combined cow and bull reference data set, with the accuracy of the phenotypes differing between the cow and bull genomic selection reference populations. The developed bivariate Bayesian stochastic search variable selection model allowed for an unbalanced design by imputing residuals in the residual updating scheme for all missing records. The performance of this model is demonstrated on a real data example, where the analyzed trait, being milk fat or protein yield, was either measured only on a cow or a bull reference population, or recorded on both. Our results were that the developed bivariate Bayesian stochastic search variable selection model was able to analyze 2 traits, even though animals had measurements on only 1 of 2 traits. The Bayesian stochastic search variable selection model yielded consistently higher accuracy for fat yield compared with a model without variable selection, both for the univariate and bivariate analyses, whereas the accuracy of both models was very similar for protein yield. The bivariate model identified several additional quantitative trait loci peaks compared with the single-trait models on either trait. In addition, the bivariate models showed a marginal increase in accuracy of genomic predictions for the cow traits (0.01-0.05), although a greater increase in accuracy is expected as the size of the bull population increases. Our results emphasize that the chosen value of priors in Bayesian genomic prediction models are especially important in small data sets. Copyright © 2013 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.

Maximum Entropy Discrimination Poisson Regression for Software Reliability Modeling.

PubMed

Chatzis, Sotirios P; Andreou, Andreas S

2015-11-01

Reliably predicting software defects is one of the most significant tasks in software engineering. Two of the major components of modern software reliability modeling approaches are: 1) extraction of salient features for software system representation, based on appropriately designed software metrics and 2) development of intricate regression models for count data, to allow effective software reliability data modeling and prediction. Surprisingly, research in the latter frontier of count data regression modeling has been rather limited. More specifically, a lack of simple and efficient algorithms for posterior computation has made the Bayesian approaches appear unattractive, and thus underdeveloped in the context of software reliability modeling. In this paper, we try to address these issues by introducing a novel Bayesian regression model for count data, based on the concept of max-margin data modeling, effected in the context of a fully Bayesian model treatment with simple and efficient posterior distribution updates. Our novel approach yields a more discriminative learning technique, making more effective use of our training data during model inference. In addition, it allows of better handling uncertainty in the modeled data, which can be a significant problem when the training data are limited. We derive elegant inference algorithms for our model under the mean-field paradigm and exhibit its effectiveness using the publicly available benchmark data sets.

Bayesian networks improve causal environmental ...

EPA Pesticide Factsheets

Rule-based weight of evidence approaches to ecological risk assessment may not account for uncertainties and generally lack probabilistic integration of lines of evidence. Bayesian networks allow causal inferences to be made from evidence by including causal knowledge about the problem, using this knowledge with probabilistic calculus to combine multiple lines of evidence, and minimizing biases in predicting or diagnosing causal relationships. Too often, sources of uncertainty in conventional weight of evidence approaches are ignored that can be accounted for with Bayesian networks. Specifying and propagating uncertainties improve the ability of models to incorporate strength of the evidence in the risk management phase of an assessment. Probabilistic inference from a Bayesian network allows evaluation of changes in uncertainty for variables from the evidence. The network structure and probabilistic framework of a Bayesian approach provide advantages over qualitative approaches in weight of evidence for capturing the impacts of multiple sources of quantifiable uncertainty on predictions of ecological risk. Bayesian networks can facilitate the development of evidence-based policy under conditions of uncertainty by incorporating analytical inaccuracies or the implications of imperfect information, structuring and communicating causal issues through qualitative directed graph formulations, and quantitatively comparing the causal power of multiple stressors on value

Error Quantification and Confidence Assessment of Aerothermal Model Predictions for Hypersonic Aircraft (Preprint)

DTIC Science & Technology

2013-09-01

based confidence metric is used to compare several different model predictions with the experimental data. II. Aerothermal Model Definition and...whereas 5% measurement uncertainty is assumed for aerodynamic pressure and heat flux measurements 4p y and 4Q y . Bayesian updating according... definitive conclusions for these particular aerodynamic models. However, given the confidence associated with the 4 sdp predictions for Run 30 (H/D

Estimation of Model's Marginal likelihood Using Adaptive Sparse Grid Surrogates in Bayesian Model Averaging

NASA Astrophysics Data System (ADS)

Zeng, X.

2015-12-01

A large number of model executions are required to obtain alternative conceptual models' predictions and their posterior probabilities in Bayesian model averaging (BMA). The posterior model probability is estimated through models' marginal likelihood and prior probability. The heavy computation burden hinders the implementation of BMA prediction, especially for the elaborated marginal likelihood estimator. For overcoming the computation burden of BMA, an adaptive sparse grid (SG) stochastic collocation method is used to build surrogates for alternative conceptual models through the numerical experiment of a synthetical groundwater model. BMA predictions depend on model posterior weights (or marginal likelihoods), and this study also evaluated four marginal likelihood estimators, including arithmetic mean estimator (AME), harmonic mean estimator (HME), stabilized harmonic mean estimator (SHME), and thermodynamic integration estimator (TIE). The results demonstrate that TIE is accurate in estimating conceptual models' marginal likelihoods. The BMA-TIE has better predictive performance than other BMA predictions. TIE has high stability for estimating conceptual model's marginal likelihood. The repeated estimated conceptual model's marginal likelihoods by TIE have significant less variability than that estimated by other estimators. In addition, the SG surrogates are efficient to facilitate BMA predictions, especially for BMA-TIE. The number of model executions needed for building surrogates is 4.13%, 6.89%, 3.44%, and 0.43% of the required model executions of BMA-AME, BMA-HME, BMA-SHME, and BMA-TIE, respectively.

Prospects and Potential Uses of Genomic Prediction of Key Performance Traits in Tetraploid Potato.

PubMed

Stich, Benjamin; Van Inghelandt, Delphine

2018-01-01

Genomic prediction is a routine tool in breeding programs of most major animal and plant species. However, its usefulness for potato breeding has not yet been evaluated in detail. The objectives of this study were to (i) examine the prospects of genomic prediction of key performance traits in a diversity panel of tetraploid potato modeling additive, dominance, and epistatic effects, (ii) investigate the effects of size and make up of training set, number of test environments and molecular markers on prediction accuracy, and (iii) assess the effect of including markers from candidate genes on the prediction accuracy. With genomic best linear unbiased prediction (GBLUP), BayesA, BayesCπ, and Bayesian LASSO, four different prediction methods were used for genomic prediction of relative area under disease progress curve after a Phytophthora infestans infection, plant maturity, maturity corrected resistance, tuber starch content, tuber starch yield (TSY), and tuber yield (TY) of 184 tetraploid potato clones or subsets thereof genotyped with the SolCAP 8.3k SNP array. The cross-validated prediction accuracies with GBLUP and the three Bayesian approaches for the six evaluated traits ranged from about 0.5 to about 0.8. For traits with a high expected genetic complexity, such as TSY and TY, we observed an 8% higher prediction accuracy using a model with additive and dominance effects compared with a model with additive effects only. Our results suggest that for oligogenic traits in general and when diagnostic markers are available in particular, the use of Bayesian methods for genomic prediction is highly recommended and that the diagnostic markers should be modeled as fixed effects. The evaluation of the relative performance of genomic prediction vs. phenotypic selection indicated that the former is superior, assuming cycle lengths and selection intensities that are possible to realize in commercial potato breeding programs.

Prospects and Potential Uses of Genomic Prediction of Key Performance Traits in Tetraploid Potato

PubMed Central

Stich, Benjamin; Van Inghelandt, Delphine

2018-01-01

Genomic prediction is a routine tool in breeding programs of most major animal and plant species. However, its usefulness for potato breeding has not yet been evaluated in detail. The objectives of this study were to (i) examine the prospects of genomic prediction of key performance traits in a diversity panel of tetraploid potato modeling additive, dominance, and epistatic effects, (ii) investigate the effects of size and make up of training set, number of test environments and molecular markers on prediction accuracy, and (iii) assess the effect of including markers from candidate genes on the prediction accuracy. With genomic best linear unbiased prediction (GBLUP), BayesA, BayesCπ, and Bayesian LASSO, four different prediction methods were used for genomic prediction of relative area under disease progress curve after a Phytophthora infestans infection, plant maturity, maturity corrected resistance, tuber starch content, tuber starch yield (TSY), and tuber yield (TY) of 184 tetraploid potato clones or subsets thereof genotyped with the SolCAP 8.3k SNP array. The cross-validated prediction accuracies with GBLUP and the three Bayesian approaches for the six evaluated traits ranged from about 0.5 to about 0.8. For traits with a high expected genetic complexity, such as TSY and TY, we observed an 8% higher prediction accuracy using a model with additive and dominance effects compared with a model with additive effects only. Our results suggest that for oligogenic traits in general and when diagnostic markers are available in particular, the use of Bayesian methods for genomic prediction is highly recommended and that the diagnostic markers should be modeled as fixed effects. The evaluation of the relative performance of genomic prediction vs. phenotypic selection indicated that the former is superior, assuming cycle lengths and selection intensities that are possible to realize in commercial potato breeding programs. PMID:29563919

Universal Darwinism As a Process of Bayesian Inference.

PubMed

Campbell, John O

2016-01-01

Many of the mathematical frameworks describing natural selection are equivalent to Bayes' Theorem, also known as Bayesian updating. By definition, a process of Bayesian Inference is one which involves a Bayesian update, so we may conclude that these frameworks describe natural selection as a process of Bayesian inference. Thus, natural selection serves as a counter example to a widely-held interpretation that restricts Bayesian Inference to human mental processes (including the endeavors of statisticians). As Bayesian inference can always be cast in terms of (variational) free energy minimization, natural selection can be viewed as comprising two components: a generative model of an "experiment" in the external world environment, and the results of that "experiment" or the "surprise" entailed by predicted and actual outcomes of the "experiment." Minimization of free energy implies that the implicit measure of "surprise" experienced serves to update the generative model in a Bayesian manner. This description closely accords with the mechanisms of generalized Darwinian process proposed both by Dawkins, in terms of replicators and vehicles, and Campbell, in terms of inferential systems. Bayesian inference is an algorithm for the accumulation of evidence-based knowledge. This algorithm is now seen to operate over a wide range of evolutionary processes, including natural selection, the evolution of mental models and cultural evolutionary processes, notably including science itself. The variational principle of free energy minimization may thus serve as a unifying mathematical framework for universal Darwinism, the study of evolutionary processes operating throughout nature.

Universal Darwinism As a Process of Bayesian Inference

PubMed Central

Campbell, John O.

2016-01-01

Many of the mathematical frameworks describing natural selection are equivalent to Bayes' Theorem, also known as Bayesian updating. By definition, a process of Bayesian Inference is one which involves a Bayesian update, so we may conclude that these frameworks describe natural selection as a process of Bayesian inference. Thus, natural selection serves as a counter example to a widely-held interpretation that restricts Bayesian Inference to human mental processes (including the endeavors of statisticians). As Bayesian inference can always be cast in terms of (variational) free energy minimization, natural selection can be viewed as comprising two components: a generative model of an “experiment” in the external world environment, and the results of that “experiment” or the “surprise” entailed by predicted and actual outcomes of the “experiment.” Minimization of free energy implies that the implicit measure of “surprise” experienced serves to update the generative model in a Bayesian manner. This description closely accords with the mechanisms of generalized Darwinian process proposed both by Dawkins, in terms of replicators and vehicles, and Campbell, in terms of inferential systems. Bayesian inference is an algorithm for the accumulation of evidence-based knowledge. This algorithm is now seen to operate over a wide range of evolutionary processes, including natural selection, the evolution of mental models and cultural evolutionary processes, notably including science itself. The variational principle of free energy minimization may thus serve as a unifying mathematical framework for universal Darwinism, the study of evolutionary processes operating throughout nature. PMID:27375438

A flexible Bayesian assessment for the expected impact of data on prediction confidence for optimal sampling designs

NASA Astrophysics Data System (ADS)

Leube, Philipp; Geiges, Andreas; Nowak, Wolfgang

2010-05-01

Incorporating hydrogeological data, such as head and tracer data, into stochastic models of subsurface flow and transport helps to reduce prediction uncertainty. Considering limited financial resources available for the data acquisition campaign, information needs towards the prediction goal should be satisfied in a efficient and task-specific manner. For finding the best one among a set of design candidates, an objective function is commonly evaluated, which measures the expected impact of data on prediction confidence, prior to their collection. An appropriate approach to this task should be stochastically rigorous, master non-linear dependencies between data, parameters and model predictions, and allow for a wide variety of different data types. Existing methods fail to fulfill all these requirements simultaneously. For this reason, we introduce a new method, denoted as CLUE (Cross-bred Likelihood Uncertainty Estimator), that derives the essential distributions and measures of data utility within a generalized, flexible and accurate framework. The method makes use of Bayesian GLUE (Generalized Likelihood Uncertainty Estimator) and extends it to an optimal design method by marginalizing over the yet unknown data values. Operating in a purely Bayesian Monte-Carlo framework, CLUE is a strictly formal information processing scheme free of linearizations. It provides full flexibility associated with the type of measurements (linear, non-linear, direct, indirect) and accounts for almost arbitrary sources of uncertainty (e.g. heterogeneity, geostatistical assumptions, boundary conditions, model concepts) via stochastic simulation and Bayesian model averaging. This helps to minimize the strength and impact of possible subjective prior assumptions, that would be hard to defend prior to data collection. Our study focuses on evaluating two different uncertainty measures: (i) expected conditional variance and (ii) expected relative entropy of a given prediction goal. The applicability and advantages are shown in a synthetic example. Therefor, we consider a contaminant source, posing a threat on a drinking water well in an aquifer. Furthermore, we assume uncertainty in geostatistical parameters, boundary conditions and hydraulic gradient. The two mentioned measures evaluate the sensitivity of (1) general prediction confidence and (2) exceedance probability of a legal regulatory threshold value on sampling locations.

Bayesian model selection: Evidence estimation based on DREAM simulation and bridge sampling

NASA Astrophysics Data System (ADS)

Volpi, Elena; Schoups, Gerrit; Firmani, Giovanni; Vrugt, Jasper A.

2017-04-01

Bayesian inference has found widespread application in Earth and Environmental Systems Modeling, providing an effective tool for prediction, data assimilation, parameter estimation, uncertainty analysis and hypothesis testing. Under multiple competing hypotheses, the Bayesian approach also provides an attractive alternative to traditional information criteria (e.g. AIC, BIC) for model selection. The key variable for Bayesian model selection is the evidence (or marginal likelihood) that is the normalizing constant in the denominator of Bayes theorem; while it is fundamental for model selection, the evidence is not required for Bayesian inference. It is computed for each hypothesis (model) by averaging the likelihood function over the prior parameter distribution, rather than maximizing it as by information criteria; the larger a model evidence the more support it receives among a collection of hypothesis as the simulated values assign relatively high probability density to the observed data. Hence, the evidence naturally acts as an Occam's razor, preferring simpler and more constrained models against the selection of over-fitted ones by information criteria that incorporate only the likelihood maximum. Since it is not particularly easy to estimate the evidence in practice, Bayesian model selection via the marginal likelihood has not yet found mainstream use. We illustrate here the properties of a new estimator of the Bayesian model evidence, which provides robust and unbiased estimates of the marginal likelihood; the method is coined Gaussian Mixture Importance Sampling (GMIS). GMIS uses multidimensional numerical integration of the posterior parameter distribution via bridge sampling (a generalization of importance sampling) of a mixture distribution fitted to samples of the posterior distribution derived from the DREAM algorithm (Vrugt et al., 2008; 2009). Some illustrative examples are presented to show the robustness and superiority of the GMIS estimator with respect to other commonly used approaches in the literature.

[Development and Application of a Performance Prediction Model for Home Care Nursing Based on a Balanced Scorecard using the Bayesian Belief Network].

PubMed

Noh, Wonjung; Seomun, Gyeongae

2015-06-01

This study was conducted to develop key performance indicators (KPIs) for home care nursing (HCN) based on a balanced scorecard, and to construct a performance prediction model of strategic objectives using the Bayesian Belief Network (BBN). This methodological study included four steps: establishment of KPIs, performance prediction modeling, development of a performance prediction model using BBN, and simulation of a suggested nursing management strategy. An HCN expert group and a staff group participated. The content validity index was analyzed using STATA 13.0, and BBN was analyzed using HUGIN 8.0. We generated a list of KPIs composed of 4 perspectives, 10 strategic objectives, and 31 KPIs. In the validity test of the performance prediction model, the factor with the greatest variance for increasing profit was maximum cost reduction of HCN services. The factor with the smallest variance for increasing profit was a minimum image improvement for HCN. During sensitivity analysis, the probability of the expert group did not affect the sensitivity. Furthermore, simulation of a 10% image improvement predicted the most effective way to increase profit. KPIs of HCN can estimate financial and non-financial performance. The performance prediction model for HCN will be useful to improve performance.

A Bayesian Approach Based Outage Prediction in Electric Utility Systems Using Radar Measurement Data

DOE PAGES

Yue, Meng; Toto, Tami; Jensen, Michael P.; ...

2017-05-18

Severe weather events such as strong thunderstorms are some of the most significant and frequent threats to the electrical grid infrastructure. Outages resulting from storms can be very costly. While some tools are available to utilities to predict storm occurrences and damage, they are typically very crude and provide little means of facilitating restoration efforts. This study developed a methodology to use historical high-resolution (both temporal and spatial) radar observations of storm characteristics and outage information to develop weather condition dependent failure rate models (FRMs) for different grid components. Such models can provide an estimation or prediction of the outagemore » numbers in small areas of a utility’s service territory once the real-time measurement or forecasted data of weather conditions become available as the input to the models. Considering the potential value provided by real-time outages reported, a Bayesian outage prediction (BOP) algorithm is proposed to account for both strength and uncertainties of the reported outages and failure rate models. The potential benefit of this outage prediction scheme is illustrated in this study.« less

A Bayesian Approach Based Outage Prediction in Electric Utility Systems Using Radar Measurement Data

DOE Office of Scientific and Technical Information (OSTI.GOV)

Yue, Meng; Toto, Tami; Jensen, Michael P.

Severe weather events such as strong thunderstorms are some of the most significant and frequent threats to the electrical grid infrastructure. Outages resulting from storms can be very costly. While some tools are available to utilities to predict storm occurrences and damage, they are typically very crude and provide little means of facilitating restoration efforts. This study developed a methodology to use historical high-resolution (both temporal and spatial) radar observations of storm characteristics and outage information to develop weather condition dependent failure rate models (FRMs) for different grid components. Such models can provide an estimation or prediction of the outagemore » numbers in small areas of a utility’s service territory once the real-time measurement or forecasted data of weather conditions become available as the input to the models. Considering the potential value provided by real-time outages reported, a Bayesian outage prediction (BOP) algorithm is proposed to account for both strength and uncertainties of the reported outages and failure rate models. The potential benefit of this outage prediction scheme is illustrated in this study.« less

Logistic Mixed Models to Investigate Implicit and Explicit Belief Tracking

PubMed Central

Lages, Martin; Scheel, Anne

2016-01-01

We investigated the proposition of a two-systems Theory of Mind in adults’ belief tracking. A sample of N = 45 participants predicted the choice of one of two opponent players after observing several rounds in an animated card game. Three matches of this card game were played and initial gaze direction on target and subsequent choice predictions were recorded for each belief task and participant. We conducted logistic regressions with mixed effects on the binary data and developed Bayesian logistic mixed models to infer implicit and explicit mentalizing in true belief and false belief tasks. Although logistic regressions with mixed effects predicted the data well a Bayesian logistic mixed model with latent task- and subject-specific parameters gave a better account of the data. As expected explicit choice predictions suggested a clear understanding of true and false beliefs (TB/FB). Surprisingly, however, model parameters for initial gaze direction also indicated belief tracking. We discuss why task-specific parameters for initial gaze directions are different from choice predictions yet reflect second-order perspective taking. PMID:27853440

A Bayesian Multilevel Model for Microcystin Prediction in ...

EPA Pesticide Factsheets

The frequency of cyanobacteria blooms in North American lakes is increasing. A major concernwith rising cyanobacteria blooms is microcystin, a common cyanobacterial hepatotoxin. Toexplore the conditions that promote high microcystin concentrations, we analyzed the US EPANational Lake Assessment (NLA) dataset collected in the summer of 2007. The NLA datasetis reported for nine eco-regions. We used the results of random forest modeling as a means ofvariable selection from which we developed a Bayesian multilevel model of microcystin concentrations.Model parameters under a multilevel modeling framework are eco-region specific, butthey are also assumed to be exchangeable across eco-regions for broad continental scaling. Theexchangeability assumption ensures that both the common patterns and eco-region specific featureswill be reflected in the model. Furthermore, the method incorporates appropriate estimatesof uncertainty. Our preliminary results show associations between microcystin and turbidity, totalnutrients, and N:P ratios. The NLA 2012 will be used for Bayesian updating. The results willhelp develop management strategies to alleviate microcystin impacts and improve lake quality. This work provides a probabilistic framework for predicting microcystin presences in lakes. It would allow for insights to be made about how changes in nutrient concentrations could potentially change toxin levels.

Sequential Designs Based on Bayesian Uncertainty Quantification in Sparse Representation Surrogate Modeling

DOE Office of Scientific and Technical Information (OSTI.GOV)

Chen, Ray -Bing; Wang, Weichung; Jeff Wu, C. F.

A numerical method, called OBSM, was recently proposed which employs overcomplete basis functions to achieve sparse representations. While the method can handle non-stationary response without the need of inverting large covariance matrices, it lacks the capability to quantify uncertainty in predictions. We address this issue by proposing a Bayesian approach which first imposes a normal prior on the large space of linear coefficients, then applies the MCMC algorithm to generate posterior samples for predictions. From these samples, Bayesian credible intervals can then be obtained to assess prediction uncertainty. A key application for the proposed method is the efficient construction ofmore » sequential designs. Several sequential design procedures with different infill criteria are proposed based on the generated posterior samples. As a result, numerical studies show that the proposed schemes are capable of solving problems of positive point identification, optimization, and surrogate fitting.« less

Sequential Designs Based on Bayesian Uncertainty Quantification in Sparse Representation Surrogate Modeling

DOE PAGES

Chen, Ray -Bing; Wang, Weichung; Jeff Wu, C. F.

2017-04-12

A numerical method, called OBSM, was recently proposed which employs overcomplete basis functions to achieve sparse representations. While the method can handle non-stationary response without the need of inverting large covariance matrices, it lacks the capability to quantify uncertainty in predictions. We address this issue by proposing a Bayesian approach which first imposes a normal prior on the large space of linear coefficients, then applies the MCMC algorithm to generate posterior samples for predictions. From these samples, Bayesian credible intervals can then be obtained to assess prediction uncertainty. A key application for the proposed method is the efficient construction ofmore » sequential designs. Several sequential design procedures with different infill criteria are proposed based on the generated posterior samples. As a result, numerical studies show that the proposed schemes are capable of solving problems of positive point identification, optimization, and surrogate fitting.« less

Bayesian transformation cure frailty models with multivariate failure time data.

PubMed

Yin, Guosheng

2008-12-10

We propose a class of transformation cure frailty models to accommodate a survival fraction in multivariate failure time data. Established through a general power transformation, this family of cure frailty models includes the proportional hazards and the proportional odds modeling structures as two special cases. Within the Bayesian paradigm, we obtain the joint posterior distribution and the corresponding full conditional distributions of the model parameters for the implementation of Gibbs sampling. Model selection is based on the conditional predictive ordinate statistic and deviance information criterion. As an illustration, we apply the proposed method to a real data set from dentistry.

Development of a clinical decision model for thyroid nodules.

PubMed

Stojadinovic, Alexander; Peoples, George E; Libutti, Steven K; Henry, Leonard R; Eberhardt, John; Howard, Robin S; Gur, David; Elster, Eric A; Nissan, Aviram

2009-08-10

Thyroid nodules represent a common problem brought to medical attention. Four to seven percent of the United States adult population (10-18 million people) has a palpable thyroid nodule, however the majority (>95%) of thyroid nodules are benign. While, fine needle aspiration remains the most cost effective and accurate diagnostic tool for thyroid nodules in current practice, over 20% of patients undergoing FNA of a thyroid nodule have indeterminate cytology (follicular neoplasm) with associated malignancy risk prevalence of 20-30%. These patients require thyroid lobectomy/isthmusectomy purely for the purpose of attaining a definitive diagnosis. Given that the majority (70-80%) of these patients have benign surgical pathology, thyroidectomy in these patients is conducted principally with diagnostic intent. Clinical models predictive of malignancy risk are needed to support treatment decisions in patients with thyroid nodules in order to reduce morbidity associated with unnecessary diagnostic surgery. Data were analyzed from a completed prospective cohort trial conducted over a 4-year period involving 216 patients with thyroid nodules undergoing ultrasound (US), electrical impedance scanning (EIS) and fine needle aspiration cytology (FNA) prior to thyroidectomy. A Bayesian model was designed to predict malignancy in thyroid nodules based on multivariate dependence relationships between independent covariates. Ten-fold cross-validation was performed to estimate classifier error wherein the data set was randomized into ten separate and unique train and test sets consisting of a training set (90% of records) and a test set (10% of records). A receiver-operating-characteristics (ROC) curve of these predictions and area under the curve (AUC) were calculated to determine model robustness for predicting malignancy in thyroid nodules. Thyroid nodule size, FNA cytology, US and EIS characteristics were highly predictive of malignancy. Cross validation of the model created with Bayesian Network Analysis effectively predicted malignancy [AUC = 0.88 (95%CI: 0.82-0.94)] in thyroid nodules. The positive and negative predictive values of the model are 83% (95%CI: 76%-91%) and 79% (95%CI: 72%-86%), respectively. An integrated predictive decision model using Bayesian inference incorporating readily obtainable thyroid nodule measures is clinically relevant, as it effectively predicts malignancy in thyroid nodules. This model warrants further validation testing in prospective clinical trials.

Bayesian spatio-temporal modeling of particulate matter concentrations in Peninsular Malaysia

NASA Astrophysics Data System (ADS)

Manga, Edna; Awang, Norhashidah

2016-06-01

This article presents an application of a Bayesian spatio-temporal Gaussian process (GP) model on particulate matter concentrations from Peninsular Malaysia. We analyze daily PM10 concentration levels from 35 monitoring sites in June and July 2011. The spatiotemporal model set in a Bayesian hierarchical framework allows for inclusion of informative covariates, meteorological variables and spatiotemporal interactions. Posterior density estimates of the model parameters are obtained by Markov chain Monte Carlo methods. Preliminary data analysis indicate information on PM10 levels at sites classified as industrial locations could explain part of the space time variations. We include the site-type indicator in our modeling efforts. Results of the parameter estimates for the fitted GP model show significant spatio-temporal structure and positive effect of the location-type explanatory variable. We also compute some validation criteria for the out of sample sites that show the adequacy of the model for predicting PM10 at unmonitored sites.

A Bayesian estimation of a stochastic predator-prey model of economic fluctuations

NASA Astrophysics Data System (ADS)

Dibeh, Ghassan; Luchinsky, Dmitry G.; Luchinskaya, Daria D.; Smelyanskiy, Vadim N.

2007-06-01

In this paper, we develop a Bayesian framework for the empirical estimation of the parameters of one of the best known nonlinear models of the business cycle: The Marx-inspired model of a growth cycle introduced by R. M. Goodwin. The model predicts a series of closed cycles representing the dynamics of labor's share and the employment rate in the capitalist economy. The Bayesian framework is used to empirically estimate a modified Goodwin model. The original model is extended in two ways. First, we allow for exogenous periodic variations of the otherwise steady growth rates of the labor force and productivity per worker. Second, we allow for stochastic variations of those parameters. The resultant modified Goodwin model is a stochastic predator-prey model with periodic forcing. The model is then estimated using a newly developed Bayesian estimation method on data sets representing growth cycles in France and Italy during the years 1960-2005. Results show that inference of the parameters of the stochastic Goodwin model can be achieved. The comparison of the dynamics of the Goodwin model with the inferred values of parameters demonstrates quantitative agreement with the growth cycle empirical data.

CRAFFT: An Activity Prediction Model based on Bayesian Networks

PubMed Central

Nazerfard, Ehsan; Cook, Diane J.

2014-01-01

Recent advances in the areas of pervasive computing, data mining, and machine learning offer unique opportunities to provide health monitoring and assistance for individuals facing difficulties to live independently in their homes. Several components have to work together to provide health monitoring for smart home residents including, but not limited to, activity recognition, activity discovery, activity prediction, and prompting system. Compared to the significant research done to discover and recognize activities, less attention has been given to predict the future activities that the resident is likely to perform. Activity prediction components can play a major role in design of a smart home. For instance, by taking advantage of an activity prediction module, a smart home can learn context-aware rules to prompt individuals to initiate important activities. In this paper, we propose an activity prediction model using Bayesian networks together with a novel two-step inference process to predict both the next activity features and the next activity label. We also propose an approach to predict the start time of the next activity which is based on modeling the relative start time of the predicted activity using the continuous normal distribution and outlier detection. To validate our proposed models, we used real data collected from physical smart environments. PMID:25937847

CRAFFT: An Activity Prediction Model based on Bayesian Networks.

PubMed

Nazerfard, Ehsan; Cook, Diane J

2015-04-01

Recent advances in the areas of pervasive computing, data mining, and machine learning offer unique opportunities to provide health monitoring and assistance for individuals facing difficulties to live independently in their homes. Several components have to work together to provide health monitoring for smart home residents including, but not limited to, activity recognition, activity discovery, activity prediction, and prompting system. Compared to the significant research done to discover and recognize activities, less attention has been given to predict the future activities that the resident is likely to perform. Activity prediction components can play a major role in design of a smart home. For instance, by taking advantage of an activity prediction module, a smart home can learn context-aware rules to prompt individuals to initiate important activities. In this paper, we propose an activity prediction model using Bayesian networks together with a novel two-step inference process to predict both the next activity features and the next activity label. We also propose an approach to predict the start time of the next activity which is based on modeling the relative start time of the predicted activity using the continuous normal distribution and outlier detection. To validate our proposed models, we used real data collected from physical smart environments.

Model based inference from microvascular measurements: Combining experimental measurements and model predictions using a Bayesian probabilistic approach

PubMed Central

Rasmussen, Peter M.; Smith, Amy F.; Sakadžić, Sava; Boas, David A.; Pries, Axel R.; Secomb, Timothy W.; Østergaard, Leif

2017-01-01

Objective In vivo imaging of the microcirculation and network-oriented modeling have emerged as powerful means of studying microvascular function and understanding its physiological significance. Network-oriented modeling may provide the means of summarizing vast amounts of data produced by high-throughput imaging techniques in terms of key, physiological indices. To estimate such indices with sufficient certainty, however, network-oriented analysis must be robust to the inevitable presence of uncertainty due to measurement errors as well as model errors. Methods We propose the Bayesian probabilistic data analysis framework as a means of integrating experimental measurements and network model simulations into a combined and statistically coherent analysis. The framework naturally handles noisy measurements and provides posterior distributions of model parameters as well as physiological indices associated with uncertainty. Results We applied the analysis framework to experimental data from three rat mesentery networks and one mouse brain cortex network. We inferred distributions for more than five hundred unknown pressure and hematocrit boundary conditions. Model predictions were consistent with previous analyses, and remained robust when measurements were omitted from model calibration. Conclusion Our Bayesian probabilistic approach may be suitable for optimizing data acquisition and for analyzing and reporting large datasets acquired as part of microvascular imaging studies. PMID:27987383

SU-F-R-44: Modeling Lung SBRT Tumor Response Using Bayesian Network Averaging

DOE Office of Scientific and Technical Information (OSTI.GOV)

Diamant, A; Ybarra, N; Seuntjens, J

2016-06-15

Purpose: The prediction of tumor control after a patient receives lung SBRT (stereotactic body radiation therapy) has proven to be challenging, due to the complex interactions between an individual’s biology and dose-volume metrics. Many of these variables have predictive power when combined, a feature that we exploit using a graph modeling approach based on Bayesian networks. This provides a probabilistic framework that allows for accurate and visually intuitive predictive modeling. The aim of this study is to uncover possible interactions between an individual patient’s characteristics and generate a robust model capable of predicting said patient’s treatment outcome. Methods: We investigatedmore » a cohort of 32 prospective patients from multiple institutions whom had received curative SBRT to the lung. The number of patients exhibiting tumor failure was observed to be 7 (event rate of 22%). The serum concentration of 5 biomarkers previously associated with NSCLC (non-small cell lung cancer) was measured pre-treatment. A total of 21 variables were analyzed including: dose-volume metrics with BED (biologically effective dose) correction and clinical variables. A Markov Chain Monte Carlo technique estimated the posterior probability distribution of the potential graphical structures. The probability of tumor failure was then estimated by averaging the top 100 graphs and applying Baye’s rule. Results: The optimal Bayesian model generated throughout this study incorporated the PTV volume, the serum concentration of the biomarker EGFR (epidermal growth factor receptor) and prescription BED. This predictive model recorded an area under the receiver operating characteristic curve of 0.94(1), providing better performance compared to competing methods in other literature. Conclusion: The use of biomarkers in conjunction with dose-volume metrics allows for the generation of a robust predictive model. The preliminary results of this report demonstrate that it is possible to accurately model the prognosis of an individual lung SBRT patient’s treatment.« less

Adaptive selection and validation of models of complex systems in the presence of uncertainty

DOE Office of Scientific and Technical Information (OSTI.GOV)

Farrell-Maupin, Kathryn; Oden, J. T.

This study describes versions of OPAL, the Occam-Plausibility Algorithm in which the use of Bayesian model plausibilities is replaced with information theoretic methods, such as the Akaike Information Criterion and the Bayes Information Criterion. Applications to complex systems of coarse-grained molecular models approximating atomistic models of polyethylene materials are described. All of these model selection methods take into account uncertainties in the model, the observational data, the model parameters, and the predicted quantities of interest. A comparison of the models chosen by Bayesian model selection criteria and those chosen by the information-theoretic criteria is given.

Adaptive selection and validation of models of complex systems in the presence of uncertainty

DOE PAGES

Farrell-Maupin, Kathryn; Oden, J. T.

2017-08-01

This study describes versions of OPAL, the Occam-Plausibility Algorithm in which the use of Bayesian model plausibilities is replaced with information theoretic methods, such as the Akaike Information Criterion and the Bayes Information Criterion. Applications to complex systems of coarse-grained molecular models approximating atomistic models of polyethylene materials are described. All of these model selection methods take into account uncertainties in the model, the observational data, the model parameters, and the predicted quantities of interest. A comparison of the models chosen by Bayesian model selection criteria and those chosen by the information-theoretic criteria is given.

Priors in Whole-Genome Regression: The Bayesian Alphabet Returns

PubMed Central

Gianola, Daniel

2013-01-01

Whole-genome enabled prediction of complex traits has received enormous attention in animal and plant breeding and is making inroads into human and even Drosophila genetics. The term “Bayesian alphabet” denotes a growing number of letters of the alphabet used to denote various Bayesian linear regressions that differ in the priors adopted, while sharing the same sampling model. We explore the role of the prior distribution in whole-genome regression models for dissecting complex traits in what is now a standard situation with genomic data where the number of unknown parameters (p) typically exceeds sample size (n). Members of the alphabet aim to confront this overparameterization in various manners, but it is shown here that the prior is always influential, unless n ≫ p. This happens because parameters are not likelihood identified, so Bayesian learning is imperfect. Since inferences are not devoid of the influence of the prior, claims about genetic architecture from these methods should be taken with caution. However, all such procedures may deliver reasonable predictions of complex traits, provided that some parameters (“tuning knobs”) are assessed via a properly conducted cross-validation. It is concluded that members of the alphabet have a room in whole-genome prediction of phenotypes, but have somewhat doubtful inferential value, at least when sample size is such that n ≪ p. PMID:23636739

Bayesian algorithm implementation in a real time exposure assessment model on benzene with calculation of associated cancer risks.

PubMed

Sarigiannis, Dimosthenis A; Karakitsios, Spyros P; Gotti, Alberto; Papaloukas, Costas L; Kassomenos, Pavlos A; Pilidis, Georgios A

2009-01-01

The objective of the current study was the development of a reliable modeling platform to calculate in real time the personal exposure and the associated health risk for filling station employees evaluating current environmental parameters (traffic, meteorological and amount of fuel traded) determined by the appropriate sensor network. A set of Artificial Neural Networks (ANNs) was developed to predict benzene exposure pattern for the filling station employees. Furthermore, a Physiology Based Pharmaco-Kinetic (PBPK) risk assessment model was developed in order to calculate the lifetime probability distribution of leukemia to the employees, fed by data obtained by the ANN model. Bayesian algorithm was involved in crucial points of both model sub compartments. The application was evaluated in two filling stations (one urban and one rural). Among several algorithms available for the development of the ANN exposure model, Bayesian regularization provided the best results and seemed to be a promising technique for prediction of the exposure pattern of that occupational population group. On assessing the estimated leukemia risk under the scope of providing a distribution curve based on the exposure levels and the different susceptibility of the population, the Bayesian algorithm was a prerequisite of the Monte Carlo approach, which is integrated in the PBPK-based risk model. In conclusion, the modeling system described herein is capable of exploiting the information collected by the environmental sensors in order to estimate in real time the personal exposure and the resulting health risk for employees of gasoline filling stations.

Bayesian Algorithm Implementation in a Real Time Exposure Assessment Model on Benzene with Calculation of Associated Cancer Risks

PubMed Central

Sarigiannis, Dimosthenis A.; Karakitsios, Spyros P.; Gotti, Alberto; Papaloukas, Costas L.; Kassomenos, Pavlos A.; Pilidis, Georgios A.

2009-01-01

The objective of the current study was the development of a reliable modeling platform to calculate in real time the personal exposure and the associated health risk for filling station employees evaluating current environmental parameters (traffic, meteorological and amount of fuel traded) determined by the appropriate sensor network. A set of Artificial Neural Networks (ANNs) was developed to predict benzene exposure pattern for the filling station employees. Furthermore, a Physiology Based Pharmaco-Kinetic (PBPK) risk assessment model was developed in order to calculate the lifetime probability distribution of leukemia to the employees, fed by data obtained by the ANN model. Bayesian algorithm was involved in crucial points of both model sub compartments. The application was evaluated in two filling stations (one urban and one rural). Among several algorithms available for the development of the ANN exposure model, Bayesian regularization provided the best results and seemed to be a promising technique for prediction of the exposure pattern of that occupational population group. On assessing the estimated leukemia risk under the scope of providing a distribution curve based on the exposure levels and the different susceptibility of the population, the Bayesian algorithm was a prerequisite of the Monte Carlo approach, which is integrated in the PBPK-based risk model. In conclusion, the modeling system described herein is capable of exploiting the information collected by the environmental sensors in order to estimate in real time the personal exposure and the resulting health risk for employees of gasoline filling stations. PMID:22399936

Application of bayesian networks to real-time flood risk estimation

NASA Astrophysics Data System (ADS)

Garrote, L.; Molina, M.; Blasco, G.

2003-04-01

This paper presents the application of a computational paradigm taken from the field of artificial intelligence - the bayesian network - to model the behaviour of hydrologic basins during floods. The final goal of this research is to develop representation techniques for hydrologic simulation models in order to define, develop and validate a mechanism, supported by a software environment, oriented to build decision models for the prediction and management of river floods in real time. The emphasis is placed on providing decision makers with tools to incorporate their knowledge of basin behaviour, usually formulated in terms of rainfall-runoff models, in the process of real-time decision making during floods. A rainfall-runoff model is only a step in the process of decision making. If a reliable rainfall forecast is available and the rainfall-runoff model is well calibrated, decisions can be based mainly on model results. However, in most practical situations, uncertainties in rainfall forecasts or model performance have to be incorporated in the decision process. The computation paradigm adopted for the simulation of hydrologic processes is the bayesian network. A bayesian network is a directed acyclic graph that represents causal influences between linked variables. Under this representation, uncertain qualitative variables are related through causal relations quantified with conditional probabilities. The solution algorithm allows the computation of the expected probability distribution of unknown variables conditioned to the observations. An approach to represent hydrologic processes by bayesian networks with temporal and spatial extensions is presented in this paper, together with a methodology for the development of bayesian models using results produced by deterministic hydrologic simulation models

Bayesian learning and the psychology of rule induction

PubMed Central

Endress, Ansgar D.

2014-01-01

In recent years, Bayesian learning models have been applied to an increasing variety of domains. While such models have been criticized on theoretical grounds, the underlying assumptions and predictions are rarely made concrete and tested experimentally. Here, I use Frank and Tenenbaum's (2011) Bayesian model of rule-learning as a case study to spell out the underlying assumptions, and to confront them with the empirical results Frank and Tenenbaum (2011) propose to simulate, as well as with novel experiments. While rule-learning is arguably well suited to rational Bayesian approaches, I show that their models are neither psychologically plausible nor ideal observer models. Further, I show that their central assumption is unfounded: humans do not always preferentially learn more specific rules, but, at least in some situations, those rules that happen to be more salient. Even when granting the unsupported assumptions, I show that all of the experiments modeled by Frank and Tenenbaum (2011) either contradict their models, or have a large number of more plausible interpretations. I provide an alternative account of the experimental data based on simple psychological mechanisms, and show that this account both describes the data better, and is easier to falsify. I conclude that, despite the recent surge in Bayesian models of cognitive phenomena, psychological phenomena are best understood by developing and testing psychological theories rather than models that can be fit to virtually any data. PMID:23454791

Optimal population prediction of sandhill crane recruitment based on climate-mediated habitat limitations.

PubMed

Gerber, Brian D; Kendall, William L; Hooten, Mevin B; Dubovsky, James A; Drewien, Roderick C

2015-09-01

1. Prediction is fundamental to scientific enquiry and application; however, ecologists tend to favour explanatory modelling. We discuss a predictive modelling framework to evaluate ecological hypotheses and to explore novel/unobserved environmental scenarios to assist conservation and management decision-makers. We apply this framework to develop an optimal predictive model for juvenile (<1 year old) sandhill crane Grus canadensis recruitment of the Rocky Mountain Population (RMP). We consider spatial climate predictors motivated by hypotheses of how drought across multiple time-scales and spring/summer weather affects recruitment. 2. Our predictive modelling framework focuses on developing a single model that includes all relevant predictor variables, regardless of collinearity. This model is then optimized for prediction by controlling model complexity using a data-driven approach that marginalizes or removes irrelevant predictors from the model. Specifically, we highlight two approaches of statistical regularization, Bayesian least absolute shrinkage and selection operator (LASSO) and ridge regression. 3. Our optimal predictive Bayesian LASSO and ridge regression models were similar and on average 37% superior in predictive accuracy to an explanatory modelling approach. Our predictive models confirmed a priori hypotheses that drought and cold summers negatively affect juvenile recruitment in the RMP. The effects of long-term drought can be alleviated by short-term wet spring-summer months; however, the alleviation of long-term drought has a much greater positive effect on juvenile recruitment. The number of freezing days and snowpack during the summer months can also negatively affect recruitment, while spring snowpack has a positive effect. 4. Breeding habitat, mediated through climate, is a limiting factor on population growth of sandhill cranes in the RMP, which could become more limiting with a changing climate (i.e. increased drought). These effects are likely not unique to cranes. The alteration of hydrological patterns and water levels by drought may impact many migratory, wetland nesting birds in the Rocky Mountains and beyond. 5. Generalizable predictive models (trained by out-of-sample fit and based on ecological hypotheses) are needed by conservation and management decision-makers. Statistical regularization improves predictions and provides a general framework for fitting models with a large number of predictors, even those with collinearity, to simultaneously identify an optimal predictive model while conducting rigorous Bayesian model selection. Our framework is important for understanding population dynamics under a changing climate and has direct applications for making harvest and habitat management decisions. Published 2015. This article is a U.S. Government work and is in the public domain in the USA.

Machine health prognostics using the Bayesian-inference-based probabilistic indication and high-order particle filtering framework

NASA Astrophysics Data System (ADS)

Yu, Jianbo

2015-12-01

Prognostics is much efficient to achieve zero-downtime performance, maximum productivity and proactive maintenance of machines. Prognostics intends to assess and predict the time evolution of machine health degradation so that machine failures can be predicted and prevented. A novel prognostics system is developed based on the data-model-fusion scheme using the Bayesian inference-based self-organizing map (SOM) and an integration of logistic regression (LR) and high-order particle filtering (HOPF). In this prognostics system, a baseline SOM is constructed to model the data distribution space of healthy machine under an assumption that predictable fault patterns are not available. Bayesian inference-based probability (BIP) derived from the baseline SOM is developed as a quantification indication of machine health degradation. BIP is capable of offering failure probability for the monitored machine, which has intuitionist explanation related to health degradation state. Based on those historic BIPs, the constructed LR and its modeling noise constitute a high-order Markov process (HOMP) to describe machine health propagation. HOPF is used to solve the HOMP estimation to predict the evolution of the machine health in the form of a probability density function (PDF). An on-line model update scheme is developed to adapt the Markov process changes to machine health dynamics quickly. The experimental results on a bearing test-bed illustrate the potential applications of the proposed system as an effective and simple tool for machine health prognostics.

Precise Network Modeling of Systems Genetics Data Using the Bayesian Network Webserver.

PubMed

Ziebarth, Jesse D; Cui, Yan

2017-01-01

The Bayesian Network Webserver (BNW, http://compbio.uthsc.edu/BNW ) is an integrated platform for Bayesian network modeling of biological datasets. It provides a web-based network modeling environment that seamlessly integrates advanced algorithms for probabilistic causal modeling and reasoning with Bayesian networks. BNW is designed for precise modeling of relatively small networks that contain less than 20 nodes. The structure learning algorithms used by BNW guarantee the discovery of the best (most probable) network structure given the data. To facilitate network modeling across multiple biological levels, BNW provides a very flexible interface that allows users to assign network nodes into different tiers and define the relationships between and within the tiers. This function is particularly useful for modeling systems genetics datasets that often consist of multiscalar heterogeneous genotype-to-phenotype data. BNW enables users to, within seconds or minutes, go from having a simply formatted input file containing a dataset to using a network model to make predictions about the interactions between variables and the potential effects of experimental interventions. In this chapter, we will introduce the functions of BNW and show how to model systems genetics datasets with BNW.

Bootstrap study of genome-enabled prediction reliabilities using haplotype blocks across Nordic Red cattle breeds.

PubMed

Cuyabano, B C D; Su, G; Rosa, G J M; Lund, M S; Gianola, D

2015-10-01

This study compared the accuracy of genome-enabled prediction models using individual single nucleotide polymorphisms (SNP) or haplotype blocks as covariates when using either a single breed or a combined population of Nordic Red cattle. The main objective was to compare predictions of breeding values of complex traits using a combined training population with haplotype blocks, with predictions using a single breed as training population and individual SNP as predictors. To compare the prediction reliabilities, bootstrap samples were taken from the test data set. With the bootstrapped samples of prediction reliabilities, we built and graphed confidence ellipses to allow comparisons. Finally, measures of statistical distances were used to calculate the gain in predictive ability. Our analyses are innovative in the context of assessment of predictive models, allowing a better understanding of prediction reliabilities and providing a statistical basis to effectively calibrate whether one prediction scenario is indeed more accurate than another. An ANOVA indicated that use of haplotype blocks produced significant gains mainly when Bayesian mixture models were used but not when Bayesian BLUP was fitted to the data. Furthermore, when haplotype blocks were used to train prediction models in a combined Nordic Red cattle population, we obtained up to a statistically significant 5.5% average gain in prediction accuracy, over predictions using individual SNP and training the model with a single breed. Copyright © 2015 American Dairy Science Association. Published by Elsevier Inc. All rights reserved.

Predictions of BuChE inhibitors using support vector machine and naive Bayesian classification techniques in drug discovery.

PubMed

Fang, Jiansong; Yang, Ranyao; Gao, Li; Zhou, Dan; Yang, Shengqian; Liu, Ai-Lin; Du, Guan-hua

2013-11-25

Butyrylcholinesterase (BuChE, EC 3.1.1.8) is an important pharmacological target for Alzheimer's disease (AD) treatment. However, the currently available BuChE inhibitor screening assays are expensive, labor-intensive, and compound-dependent. It is necessary to develop robust in silico methods to predict the activities of BuChE inhibitors for the lead identification. In this investigation, support vector machine (SVM) models and naive Bayesian models were built to discriminate BuChE inhibitors (BuChEIs) from the noninhibitors. Each molecule was initially represented in 1870 structural descriptors (1235 from ADRIANA.Code, 334 from MOE, and 301 from Discovery studio). Correlation analysis and stepwise variable selection method were applied to figure out activity-related descriptors for prediction models. Additionally, structural fingerprint descriptors were added to improve the predictive ability of models, which were measured by cross-validation, a test set validation with 1001 compounds and an external test set validation with 317 diverse chemicals. The best two models gave Matthews correlation coefficient of 0.9551 and 0.9550 for the test set and 0.9132 and 0.9221 for the external test set. To demonstrate the practical applicability of the models in virtual screening, we screened an in-house data set with 3601 compounds, and 30 compounds were selected for further bioactivity assay. The assay results showed that 10 out of 30 compounds exerted significant BuChE inhibitory activities with IC50 values ranging from 0.32 to 22.22 μM, at which three new scaffolds as BuChE inhibitors were identified for the first time. To our best knowledge, this is the first report on BuChE inhibitors using machine learning approaches. The models generated from SVM and naive Bayesian approaches successfully predicted BuChE inhibitors. The study proved the feasibility of a new method for predicting bioactivities of ligands and discovering novel lead compounds.

Bayesian Networks Improve Causal Environmental Assessments for Evidence-Based Policy.

PubMed

Carriger, John F; Barron, Mace G; Newman, Michael C

2016-12-20

Rule-based weight of evidence approaches to ecological risk assessment may not account for uncertainties and generally lack probabilistic integration of lines of evidence. Bayesian networks allow causal inferences to be made from evidence by including causal knowledge about the problem, using this knowledge with probabilistic calculus to combine multiple lines of evidence, and minimizing biases in predicting or diagnosing causal relationships. Too often, sources of uncertainty in conventional weight of evidence approaches are ignored that can be accounted for with Bayesian networks. Specifying and propagating uncertainties improve the ability of models to incorporate strength of the evidence in the risk management phase of an assessment. Probabilistic inference from a Bayesian network allows evaluation of changes in uncertainty for variables from the evidence. The network structure and probabilistic framework of a Bayesian approach provide advantages over qualitative approaches in weight of evidence for capturing the impacts of multiple sources of quantifiable uncertainty on predictions of ecological risk. Bayesian networks can facilitate the development of evidence-based policy under conditions of uncertainty by incorporating analytical inaccuracies or the implications of imperfect information, structuring and communicating causal issues through qualitative directed graph formulations, and quantitatively comparing the causal power of multiple stressors on valued ecological resources. These aspects are demonstrated through hypothetical problem scenarios that explore some major benefits of using Bayesian networks for reasoning and making inferences in evidence-based policy.

LiDAR based prediction of forest biomass using hierarchical models with spatially varying coefficients

USGS Publications Warehouse

Babcock, Chad; Finley, Andrew O.; Bradford, John B.; Kolka, Randall K.; Birdsey, Richard A.; Ryan, Michael G.

2015-01-01

Many studies and production inventory systems have shown the utility of coupling covariates derived from Light Detection and Ranging (LiDAR) data with forest variables measured on georeferenced inventory plots through regression models. The objective of this study was to propose and assess the use of a Bayesian hierarchical modeling framework that accommodates both residual spatial dependence and non-stationarity of model covariates through the introduction of spatial random effects. We explored this objective using four forest inventory datasets that are part of the North American Carbon Program, each comprising point-referenced measures of above-ground forest biomass and discrete LiDAR. For each dataset, we considered at least five regression model specifications of varying complexity. Models were assessed based on goodness of fit criteria and predictive performance using a 10-fold cross-validation procedure. Results showed that the addition of spatial random effects to the regression model intercept improved fit and predictive performance in the presence of substantial residual spatial dependence. Additionally, in some cases, allowing either some or all regression slope parameters to vary spatially, via the addition of spatial random effects, further improved model fit and predictive performance. In other instances, models showed improved fit but decreased predictive performance—indicating over-fitting and underscoring the need for cross-validation to assess predictive ability. The proposed Bayesian modeling framework provided access to pixel-level posterior predictive distributions that were useful for uncertainty mapping, diagnosing spatial extrapolation issues, revealing missing model covariates, and discovering locally significant parameters.

Application of Multiple Imputation for Missing Values in Three-Way Three-Mode Multi-Environment Trial Data

PubMed Central

Tian, Ting; McLachlan, Geoffrey J.; Dieters, Mark J.; Basford, Kaye E.

2015-01-01

It is a common occurrence in plant breeding programs to observe missing values in three-way three-mode multi-environment trial (MET) data. We proposed modifications of models for estimating missing observations for these data arrays, and developed a novel approach in terms of hierarchical clustering. Multiple imputation (MI) was used in four ways, multiple agglomerative hierarchical clustering, normal distribution model, normal regression model, and predictive mean match. The later three models used both Bayesian analysis and non-Bayesian analysis, while the first approach used a clustering procedure with randomly selected attributes and assigned real values from the nearest neighbour to the one with missing observations. Different proportions of data entries in six complete datasets were randomly selected to be missing and the MI methods were compared based on the efficiency and accuracy of estimating those values. The results indicated that the models using Bayesian analysis had slightly higher accuracy of estimation performance than those using non-Bayesian analysis but they were more time-consuming. However, the novel approach of multiple agglomerative hierarchical clustering demonstrated the overall best performances. PMID:26689369

Application of Multiple Imputation for Missing Values in Three-Way Three-Mode Multi-Environment Trial Data.

PubMed

Tian, Ting; McLachlan, Geoffrey J; Dieters, Mark J; Basford, Kaye E

2015-01-01

It is a common occurrence in plant breeding programs to observe missing values in three-way three-mode multi-environment trial (MET) data. We proposed modifications of models for estimating missing observations for these data arrays, and developed a novel approach in terms of hierarchical clustering. Multiple imputation (MI) was used in four ways, multiple agglomerative hierarchical clustering, normal distribution model, normal regression model, and predictive mean match. The later three models used both Bayesian analysis and non-Bayesian analysis, while the first approach used a clustering procedure with randomly selected attributes and assigned real values from the nearest neighbour to the one with missing observations. Different proportions of data entries in six complete datasets were randomly selected to be missing and the MI methods were compared based on the efficiency and accuracy of estimating those values. The results indicated that the models using Bayesian analysis had slightly higher accuracy of estimation performance than those using non-Bayesian analysis but they were more time-consuming. However, the novel approach of multiple agglomerative hierarchical clustering demonstrated the overall best performances.

Model inversion via multi-fidelity Bayesian optimization: a new paradigm for parameter estimation in haemodynamics, and beyond.

PubMed

Perdikaris, Paris; Karniadakis, George Em

2016-05-01

We present a computational framework for model inversion based on multi-fidelity information fusion and Bayesian optimization. The proposed methodology targets the accurate construction of response surfaces in parameter space, and the efficient pursuit to identify global optima while keeping the number of expensive function evaluations at a minimum. We train families of correlated surrogates on available data using Gaussian processes and auto-regressive stochastic schemes, and exploit the resulting predictive posterior distributions within a Bayesian optimization setting. This enables a smart adaptive sampling procedure that uses the predictive posterior variance to balance the exploration versus exploitation trade-off, and is a key enabler for practical computations under limited budgets. The effectiveness of the proposed framework is tested on three parameter estimation problems. The first two involve the calibration of outflow boundary conditions of blood flow simulations in arterial bifurcations using multi-fidelity realizations of one- and three-dimensional models, whereas the last one aims to identify the forcing term that generated a particular solution to an elliptic partial differential equation. © 2016 The Author(s).

Model inversion via multi-fidelity Bayesian optimization: a new paradigm for parameter estimation in haemodynamics, and beyond

PubMed Central

Perdikaris, Paris; Karniadakis, George Em

2016-01-01

We present a computational framework for model inversion based on multi-fidelity information fusion and Bayesian optimization. The proposed methodology targets the accurate construction of response surfaces in parameter space, and the efficient pursuit to identify global optima while keeping the number of expensive function evaluations at a minimum. We train families of correlated surrogates on available data using Gaussian processes and auto-regressive stochastic schemes, and exploit the resulting predictive posterior distributions within a Bayesian optimization setting. This enables a smart adaptive sampling procedure that uses the predictive posterior variance to balance the exploration versus exploitation trade-off, and is a key enabler for practical computations under limited budgets. The effectiveness of the proposed framework is tested on three parameter estimation problems. The first two involve the calibration of outflow boundary conditions of blood flow simulations in arterial bifurcations using multi-fidelity realizations of one- and three-dimensional models, whereas the last one aims to identify the forcing term that generated a particular solution to an elliptic partial differential equation. PMID:27194481

Research on Nonlinear Time Series Forecasting of Time-Delay NN Embedded with Bayesian Regularization

NASA Astrophysics Data System (ADS)

Jiang, Weijin; Xu, Yusheng; Xu, Yuhui; Wang, Jianmin

Based on the idea of nonlinear prediction of phase space reconstruction, this paper presented a time delay BP neural network model, whose generalization capability was improved by Bayesian regularization. Furthermore, the model is applied to forecast the imp&exp trades in one industry. The results showed that the improved model has excellent generalization capabilities, which not only learned the historical curve, but efficiently predicted the trend of business. Comparing with common evaluation of forecasts, we put on a conclusion that nonlinear forecast can not only focus on data combination and precision improvement, it also can vividly reflect the nonlinear characteristic of the forecasting system. While analyzing the forecasting precision of the model, we give a model judgment by calculating the nonlinear characteristic value of the combined serial and original serial, proved that the forecasting model can reasonably 'catch' the dynamic characteristic of the nonlinear system which produced the origin serial.

Analyzing degradation data with a random effects spline regression model

DOE PAGES

Fugate, Michael Lynn; Hamada, Michael Scott; Weaver, Brian Phillip

2017-03-17

This study proposes using a random effects spline regression model to analyze degradation data. Spline regression avoids having to specify a parametric function for the true degradation of an item. A distribution for the spline regression coefficients captures the variation of the true degradation curves from item to item. We illustrate the proposed methodology with a real example using a Bayesian approach. The Bayesian approach allows prediction of degradation of a population over time and estimation of reliability is easy to perform.

Analyzing degradation data with a random effects spline regression model

DOE Office of Scientific and Technical Information (OSTI.GOV)

Fugate, Michael Lynn; Hamada, Michael Scott; Weaver, Brian Phillip

This study proposes using a random effects spline regression model to analyze degradation data. Spline regression avoids having to specify a parametric function for the true degradation of an item. A distribution for the spline regression coefficients captures the variation of the true degradation curves from item to item. We illustrate the proposed methodology with a real example using a Bayesian approach. The Bayesian approach allows prediction of degradation of a population over time and estimation of reliability is easy to perform.

Aminoglycoside Therapy Manager: An Advanced Computer Program for Decision Support for Drug Dosing and Therapeutic Monitoring

PubMed Central

Lenert, Leslie; Lurie, Jon; Coleman, Robert; Klosterman, Heidrun; Blaschke, Terrence

1990-01-01

In this paper, we will describe an advanced drug dosing program, Aminoglycoside Therapy Manager that reasons using Bayesian pharmacokinetic modeling and symbolic modeling of patient status and drug response. Our design is similar to the design of the Digitalis Therapy Advisor program, but extends previous work by incorporating a Bayesian pharmacokinetic model, a “meta-level” analysis of drug concentrations to identify sampling errors and changes in pharmacokinetics, and including the results of the “meta-level” analysis in reasoning for dosing and therapeutic monitoring recommendations. The program is user friendly and runs on low cost general-purpose hardware. Validation studies show that the program is as accurate in predicting future drug concentrations as an expert using commercial Bayesian forecasting software.

Online Bayesian Learning with Natural Sequential Prior Distribution Used for Wind Speed Prediction

NASA Astrophysics Data System (ADS)

Cheggaga, Nawal

2017-11-01

Predicting wind speed is one of the most important and critic tasks in a wind farm. All approaches, which directly describe the stochastic dynamics of the meteorological data are facing problems related to the nature of its non-Gaussian statistics and the presence of seasonal effects .In this paper, Online Bayesian learning has been successfully applied to online learning for three-layer perceptron's used for wind speed prediction. First a conventional transition model based on the squared norm of the difference between the current parameter vector and the previous parameter vector has been used. We noticed that the transition model does not adequately consider the difference between the current and the previous wind speed measurement. To adequately consider this difference, we use a natural sequential prior. The proposed transition model uses a Fisher information matrix to consider the difference between the observation models more naturally. The obtained results showed a good agreement between both series, measured and predicted. The mean relative error over the whole data set is not exceeding 5 %.

Bayesian Optimization for Neuroimaging Pre-processing in Brain Age Classification and Prediction

PubMed Central

Lancaster, Jenessa; Lorenz, Romy; Leech, Rob; Cole, James H.

2018-01-01

Neuroimaging-based age prediction using machine learning is proposed as a biomarker of brain aging, relating to cognitive performance, health outcomes and progression of neurodegenerative disease. However, even leading age-prediction algorithms contain measurement error, motivating efforts to improve experimental pipelines. T1-weighted MRI is commonly used for age prediction, and the pre-processing of these scans involves normalization to a common template and resampling to a common voxel size, followed by spatial smoothing. Resampling parameters are often selected arbitrarily. Here, we sought to improve brain-age prediction accuracy by optimizing resampling parameters using Bayesian optimization. Using data on N = 2003 healthy individuals (aged 16–90 years) we trained support vector machines to (i) distinguish between young (<22 years) and old (>50 years) brains (classification) and (ii) predict chronological age (regression). We also evaluated generalisability of the age-regression model to an independent dataset (CamCAN, N = 648, aged 18–88 years). Bayesian optimization was used to identify optimal voxel size and smoothing kernel size for each task. This procedure adaptively samples the parameter space to evaluate accuracy across a range of possible parameters, using independent sub-samples to iteratively assess different parameter combinations to arrive at optimal values. When distinguishing between young and old brains a classification accuracy of 88.1% was achieved, (optimal voxel size = 11.5 mm3, smoothing kernel = 2.3 mm). For predicting chronological age, a mean absolute error (MAE) of 5.08 years was achieved, (optimal voxel size = 3.73 mm3, smoothing kernel = 3.68 mm). This was compared to performance using default values of 1.5 mm3 and 4mm respectively, resulting in MAE = 5.48 years, though this 7.3% improvement was not statistically significant. When assessing generalisability, best performance was achieved when applying the entire Bayesian optimization framework to the new dataset, out-performing the parameters optimized for the initial training dataset. Our study outlines the proof-of-principle that neuroimaging models for brain-age prediction can use Bayesian optimization to derive case-specific pre-processing parameters. Our results suggest that different pre-processing parameters are selected when optimization is conducted in specific contexts. This potentially motivates use of optimization techniques at many different points during the experimental process, which may improve statistical sensitivity and reduce opportunities for experimenter-led bias. PMID:29483870

Lip-reading aids word recognition most in moderate noise: a Bayesian explanation using high-dimensional feature space.

PubMed

Ma, Wei Ji; Zhou, Xiang; Ross, Lars A; Foxe, John J; Parra, Lucas C

2009-01-01

Watching a speaker's facial movements can dramatically enhance our ability to comprehend words, especially in noisy environments. From a general doctrine of combining information from different sensory modalities (the principle of inverse effectiveness), one would expect that the visual signals would be most effective at the highest levels of auditory noise. In contrast, we find, in accord with a recent paper, that visual information improves performance more at intermediate levels of auditory noise than at the highest levels, and we show that a novel visual stimulus containing only temporal information does the same. We present a Bayesian model of optimal cue integration that can explain these conflicts. In this model, words are regarded as points in a multidimensional space and word recognition is a probabilistic inference process. When the dimensionality of the feature space is low, the Bayesian model predicts inverse effectiveness; when the dimensionality is high, the enhancement is maximal at intermediate auditory noise levels. When the auditory and visual stimuli differ slightly in high noise, the model makes a counterintuitive prediction: as sound quality increases, the proportion of reported words corresponding to the visual stimulus should first increase and then decrease. We confirm this prediction in a behavioral experiment. We conclude that auditory-visual speech perception obeys the same notion of optimality previously observed only for simple multisensory stimuli.

Uncertainty quantification in capacitive RF MEMS switches

NASA Astrophysics Data System (ADS)

Pax, Benjamin J.

Development of radio frequency micro electrical-mechanical systems (RF MEMS) has led to novel approaches to implement electrical circuitry. The introduction of capacitive MEMS switches, in particular, has shown promise in low-loss, low-power devices. However, the promise of MEMS switches has not yet been completely realized. RF-MEMS switches are known to fail after only a few months of operation, and nominally similar designs show wide variability in lifetime. Modeling switch operation using nominal or as-designed parameters cannot predict the statistical spread in the number of cycles to failure, and probabilistic methods are necessary. A Bayesian framework for calibration, validation and prediction offers an integrated approach to quantifying the uncertainty in predictions of MEMS switch performance. The objective of this thesis is to use the Bayesian framework to predict the creep-related deflection of the PRISM RF-MEMS switch over several thousand hours of operation. The PRISM switch used in this thesis is the focus of research at Purdue's PRISM center, and is a capacitive contacting RF-MEMS switch. It employs a fixed-fixed nickel membrane which is electrostatically actuated by applying voltage between the membrane and a pull-down electrode. Creep plays a central role in the reliability of this switch. The focus of this thesis is on the creep model, which is calibrated against experimental data measured for a frog-leg varactor fabricated and characterized at Purdue University. Creep plasticity is modeled using plate element theory with electrostatic forces being generated using either parallel plate approximations where appropriate, or solving for the full 3D potential field. For the latter, structure-electrostatics interaction is determined through immersed boundary method. A probabilistic framework using generalized polynomial chaos (gPC) is used to create surrogate models to mitigate the costly full physics simulations, and Bayesian calibration and forward propagation of uncertainty are performed using this surrogate model. The first step in the analysis is Bayesian calibration of the creep related parameters. A computational model of the frog-leg varactor is created, and the computed creep deflection of the device over 800 hours is used to generate a surrogate model using a polynomial chaos expansion in Hermite polynomials. Parameters related to the creep phenomenon are calibrated using Bayesian calibration with experimental deflection data from the frog-leg device. The calibrated input distributions are subsequently propagated through a surrogate gPC model for the PRISM MEMS switch to produce probability density functions of the maximum membrane deflection of the membrane over several thousand hours. The assumptions related to the Bayesian calibration and forward propagation are analyzed to determine the sensitivity to these assumptions of the calibrated input distributions and propagated output distributions of the PRISM device. The work is an early step in understanding the role of geometric variability, model uncertainty, numerical errors and experimental uncertainties in the long-term performance of RF-MEMS.

The evaluation of different forest structural indices to predict the stand aboveground biomass of even-aged Scotch pine (Pinus sylvestris L.) forests in Kunduz, Northern Turkey.

PubMed

Ercanli, İlker; Kahriman, Aydın

2015-03-01

We assessed the effect of stand structural diversity, including the Shannon, improved Shannon, Simpson, McIntosh, Margelef, and Berger-Parker indices, on stand aboveground biomass (AGB) and developed statistical prediction models for the stand AGB values, including stand structural diversity indices and some stand attributes. The AGB prediction model, including only stand attributes, accounted for 85 % of the total variance in AGB (R (2)) with an Akaike's information criterion (AIC) of 807.2407, Bayesian information criterion (BIC) of 809.5397, Schwarz Bayesian criterion (SBC) of 818.0426, and root mean square error (RMSE) of 38.529 Mg. After inclusion of the stand structural diversity into the model structure, considerable improvement was observed in statistical accuracy, including 97.5 % of the total variance in AGB, with an AIC of 614.1819, BIC of 617.1242, SBC of 633.0853, and RMSE of 15.8153 Mg. The predictive fitting results indicate that some indices describing the stand structural diversity can be employed as significant independent variables to predict the AGB production of the Scotch pine stand. Further, including the stand diversity indices in the AGB prediction model with the stand attributes provided important predictive contributions in estimating the total variance in AGB.

Risk of fire occurrence in arid and semi-arid ecosystems of Iran: an investigation using Bayesian belief networks.

PubMed

Bashari, Hossein; Naghipour, Ali Asghar; Khajeddin, Seyed Jamaleddin; Sangoony, Hamed; Tahmasebi, Pejman

2016-09-01

Identifying areas that have a high risk of burning is a main component of fire management planning. Although the available tools can predict the fire risks, these are poor in accommodating uncertainties in their predictions. In this study, we accommodated uncertainty in wildfire prediction using Bayesian belief networks (BBNs). An influence diagram was developed to identify the factors influencing wildfire in arid and semi-arid areas of Iran, and it was populated with probabilities to produce a BBNs model. The behavior of the model was tested using scenario and sensitivity analysis. Land cover/use, mean annual rainfall, mean annual temperature, elevation, and livestock density were recognized as the main variables determining wildfire occurrence. The produced model had good accuracy as its ROC area under the curve was 0.986. The model could be applied in both predictive and diagnostic analysis for answering "what if" and "how" questions. The probabilistic relationships within the model can be updated over time using observation and monitoring data. The wildfire BBN model may be updated as new knowledge emerges; hence, it can be used to support the process of adaptive management.

Bayesian analysis of stochastic volatility-in-mean model with leverage and asymmetrically heavy-tailed error using generalized hyperbolic skew Student’s t-distribution*

PubMed Central

Leão, William L.; Chen, Ming-Hui

2017-01-01

A stochastic volatility-in-mean model with correlated errors using the generalized hyperbolic skew Student-t (GHST) distribution provides a robust alternative to the parameter estimation for daily stock returns in the absence of normality. An efficient Markov chain Monte Carlo (MCMC) sampling algorithm is developed for parameter estimation. The deviance information, the Bayesian predictive information and the log-predictive score criterion are used to assess the fit of the proposed model. The proposed method is applied to an analysis of the daily stock return data from the Standard & Poor’s 500 index (S&P 500). The empirical results reveal that the stochastic volatility-in-mean model with correlated errors and GH-ST distribution leads to a significant improvement in the goodness-of-fit for the S&P 500 index returns dataset over the usual normal model. PMID:29333210

Bayesian analysis of stochastic volatility-in-mean model with leverage and asymmetrically heavy-tailed error using generalized hyperbolic skew Student's t-distribution.

PubMed

Leão, William L; Abanto-Valle, Carlos A; Chen, Ming-Hui

2017-01-01

A stochastic volatility-in-mean model with correlated errors using the generalized hyperbolic skew Student-t (GHST) distribution provides a robust alternative to the parameter estimation for daily stock returns in the absence of normality. An efficient Markov chain Monte Carlo (MCMC) sampling algorithm is developed for parameter estimation. The deviance information, the Bayesian predictive information and the log-predictive score criterion are used to assess the fit of the proposed model. The proposed method is applied to an analysis of the daily stock return data from the Standard & Poor's 500 index (S&P 500). The empirical results reveal that the stochastic volatility-in-mean model with correlated errors and GH-ST distribution leads to a significant improvement in the goodness-of-fit for the S&P 500 index returns dataset over the usual normal model.

The Lifecycle of Bayesian Network Models Developed for Multi-Source Signature Assessment of Nuclear Programs

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gastelum, Zoe N.; White, Amanda M.; Whitney, Paul D.

2013-06-04

The Multi-Source Signatures for Nuclear Programs project, part of Pacific Northwest National Laboratory’s (PNNL) Signature Discovery Initiative, seeks to computationally capture expert assessment of multi-type information such as text, sensor output, imagery, or audio/video files, to assess nuclear activities through a series of Bayesian network (BN) models. These models incorporate knowledge from a diverse range of information sources in order to help assess a country’s nuclear activities. The models span engineering topic areas, state-level indicators, and facility-specific characteristics. To illustrate the development, calibration, and use of BN models for multi-source assessment, we present a model that predicts a country’s likelihoodmore » to participate in the international nuclear nonproliferation regime. We validate this model by examining the extent to which the model assists non-experts arrive at conclusions similar to those provided by nuclear proliferation experts. We also describe the PNNL-developed software used throughout the lifecycle of the Bayesian network model development.« less

A Bayesian Multilevel Model for Microcystin Prediction in ...

EPA Pesticide Factsheets

The frequency of cyanobacteria blooms in North American lakes is increasing. A major concern with rising cyanobacteria blooms is microcystin, a common cyanobacterial hepatotoxin. To explore the conditions that promote high microcystin concentrations, we analyzed the US EPA National Lake Assessment (NLA) dataset collected in the summer of 2007. The NLA dataset is reported for nine eco-regions. We used the results of random forest modeling as a means ofvariable selection from which we developed a Bayesian multilevel model of microcystin concentrations. Model parameters under a multilevel modeling framework are eco-region specific, butthey are also assumed to be exchangeable across eco-regions for broad continental scaling. The exchangeability assumption ensures that both the common patterns and eco-region specific features will be reflected in the model. Furthermore, the method incorporates appropriate estimates of uncertainty. Our preliminary results show associations between microcystin and turbidity, total nutrients, and N:P ratios. Upon release of a comparable 2012 NLA dataset, we will apply Bayesian updating. The results will help develop management strategies to alleviate microcystin impacts and improve lake quality. This work provides a probabilistic framework for predicting microcystin presences in lakes. It would allow for insights to be made about how changes in nutrient concentrations could potentially change toxin levels.

Quantum-Like Bayesian Networks for Modeling Decision Making

PubMed Central

Moreira, Catarina; Wichert, Andreas

2016-01-01

In this work, we explore an alternative quantum structure to perform quantum probabilistic inferences to accommodate the paradoxical findings of the Sure Thing Principle. We propose a Quantum-Like Bayesian Network, which consists in replacing classical probabilities by quantum probability amplitudes. However, since this approach suffers from the problem of exponential growth of quantum parameters, we also propose a similarity heuristic that automatically fits quantum parameters through vector similarities. This makes the proposed model general and predictive in contrast to the current state of the art models, which cannot be generalized for more complex decision scenarios and that only provide an explanatory nature for the observed paradoxes. In the end, the model that we propose consists in a nonparametric method for estimating inference effects from a statistical point of view. It is a statistical model that is simpler than the previous quantum dynamic and quantum-like models proposed in the literature. We tested the proposed network with several empirical data from the literature, mainly from the Prisoner's Dilemma game and the Two Stage Gambling game. The results obtained show that the proposed quantum Bayesian Network is a general method that can accommodate violations of the laws of classical probability theory and make accurate predictions regarding human decision-making in these scenarios. PMID:26858669

Model averaging, optimal inference, and habit formation

PubMed Central

FitzGerald, Thomas H. B.; Dolan, Raymond J.; Friston, Karl J.

2014-01-01

Postulating that the brain performs approximate Bayesian inference generates principled and empirically testable models of neuronal function—the subject of much current interest in neuroscience and related disciplines. Current formulations address inference and learning under some assumed and particular model. In reality, organisms are often faced with an additional challenge—that of determining which model or models of their environment are the best for guiding behavior. Bayesian model averaging—which says that an agent should weight the predictions of different models according to their evidence—provides a principled way to solve this problem. Importantly, because model evidence is determined by both the accuracy and complexity of the model, optimal inference requires that these be traded off against one another. This means an agent's behavior should show an equivalent balance. We hypothesize that Bayesian model averaging plays an important role in cognition, given that it is both optimal and realizable within a plausible neuronal architecture. We outline model averaging and how it might be implemented, and then explore a number of implications for brain and behavior. In particular, we propose that model averaging can explain a number of apparently suboptimal phenomena within the framework of approximate (bounded) Bayesian inference, focusing particularly upon the relationship between goal-directed and habitual behavior. PMID:25018724

The utility of Bayesian predictive probabilities for interim monitoring of clinical trials

PubMed Central

Connor, Jason T.; Ayers, Gregory D; Alvarez, JoAnn

2014-01-01

Background Bayesian predictive probabilities can be used for interim monitoring of clinical trials to estimate the probability of observing a statistically significant treatment effect if the trial were to continue to its predefined maximum sample size. Purpose We explore settings in which Bayesian predictive probabilities are advantageous for interim monitoring compared to Bayesian posterior probabilities, p-values, conditional power, or group sequential methods. Results For interim analyses that address prediction hypotheses, such as futility monitoring and efficacy monitoring with lagged outcomes, only predictive probabilities properly account for the amount of data remaining to be observed in a clinical trial and have the flexibility to incorporate additional information via auxiliary variables. Limitations Computational burdens limit the feasibility of predictive probabilities in many clinical trial settings. The specification of prior distributions brings additional challenges for regulatory approval. Conclusions The use of Bayesian predictive probabilities enables the choice of logical interim stopping rules that closely align with the clinical decision making process. PMID:24872363

Integrating System Dynamics and Bayesian Networks with Application to Counter-IED Scenarios

DOE Office of Scientific and Technical Information (OSTI.GOV)

Jarman, Kenneth D.; Brothers, Alan J.; Whitney, Paul D.

2010-06-06

The practice of choosing a single modeling paradigm for predictive analysis can limit the scope and relevance of predictions and their utility to decision-making processes. Considering multiple modeling methods simultaneously may improve this situation, but a better solution provides a framework for directly integrating different, potentially complementary modeling paradigms to enable more comprehensive modeling and predictions, and thus better-informed decisions. The primary challenges of this kind of model integration are to bridge language and conceptual gaps between modeling paradigms, and to determine whether natural and useful linkages can be made in a formal mathematical manner. To address these challenges inmore » the context of two specific modeling paradigms, we explore mathematical and computational options for linking System Dynamics (SD) and Bayesian network (BN) models and incorporating data into the integrated models. We demonstrate that integrated SD/BN models can naturally be described as either state space equations or Dynamic Bayes Nets, which enables the use of many existing computational methods for simulation and data integration. To demonstrate, we apply our model integration approach to techno-social models of insurgent-led attacks and security force counter-measures centered on improvised explosive devices.« less

Model Selection Methods for Mixture Dichotomous IRT Models

ERIC Educational Resources Information Center

Li, Feiming; Cohen, Allan S.; Kim, Seock-Ho; Cho, Sun-Joo

2009-01-01

This study examines model selection indices for use with dichotomous mixture item response theory (IRT) models. Five indices are considered: Akaike's information coefficient (AIC), Bayesian information coefficient (BIC), deviance information coefficient (DIC), pseudo-Bayes factor (PsBF), and posterior predictive model checks (PPMC). The five…

Bayesian Modeling of Exposure and Airflow Using Two-Zone Models

PubMed Central

Zhang, Yufen; Banerjee, Sudipto; Yang, Rui; Lungu, Claudiu; Ramachandran, Gurumurthy

2009-01-01

Mathematical modeling is being increasingly used as a means for assessing occupational exposures. However, predicting exposure in real settings is constrained by lack of quantitative knowledge of exposure determinants. Validation of models in occupational settings is, therefore, a challenge. Not only do the model parameters need to be known, the models also need to predict the output with some degree of accuracy. In this paper, a Bayesian statistical framework is used for estimating model parameters and exposure concentrations for a two-zone model. The model predicts concentrations in a zone near the source and far away from the source as functions of the toluene generation rate, air ventilation rate through the chamber, and the airflow between near and far fields. The framework combines prior or expert information on the physical model along with the observed data. The framework is applied to simulated data as well as data obtained from the experiments conducted in a chamber. Toluene vapors are generated from a source under different conditions of airflow direction, the presence of a mannequin, and simulated body heat of the mannequin. The Bayesian framework accounts for uncertainty in measurement as well as in the unknown rate of airflow between the near and far fields. The results show that estimates of the interzonal airflow are always close to the estimated equilibrium solutions, which implies that the method works efficiently. The predictions of near-field concentration for both the simulated and real data show nice concordance with the true values, indicating that the two-zone model assumptions agree with the reality to a large extent and the model is suitable for predicting the contaminant concentration. Comparison of the estimated model and its margin of error with the experimental data thus enables validation of the physical model assumptions. The approach illustrates how exposure models and information on model parameters together with the knowledge of uncertainty and variability in these quantities can be used to not only provide better estimates of model outputs but also model parameters. PMID:19403840

Bayesian assessment of the expected data impact on prediction confidence in optimal sampling design

NASA Astrophysics Data System (ADS)

Leube, P. C.; Geiges, A.; Nowak, W.

2012-02-01

Incorporating hydro(geo)logical data, such as head and tracer data, into stochastic models of (subsurface) flow and transport helps to reduce prediction uncertainty. Because of financial limitations for investigation campaigns, information needs toward modeling or prediction goals should be satisfied efficiently and rationally. Optimal design techniques find the best one among a set of investigation strategies. They optimize the expected impact of data on prediction confidence or related objectives prior to data collection. We introduce a new optimal design method, called PreDIA(gnosis) (Preposterior Data Impact Assessor). PreDIA derives the relevant probability distributions and measures of data utility within a fully Bayesian, generalized, flexible, and accurate framework. It extends the bootstrap filter (BF) and related frameworks to optimal design by marginalizing utility measures over the yet unknown data values. PreDIA is a strictly formal information-processing scheme free of linearizations. It works with arbitrary simulation tools, provides full flexibility concerning measurement types (linear, nonlinear, direct, indirect), allows for any desired task-driven formulations, and can account for various sources of uncertainty (e.g., heterogeneity, geostatistical assumptions, boundary conditions, measurement values, model structure uncertainty, a large class of model errors) via Bayesian geostatistics and model averaging. Existing methods fail to simultaneously provide these crucial advantages, which our method buys at relatively higher-computational costs. We demonstrate the applicability and advantages of PreDIA over conventional linearized methods in a synthetic example of subsurface transport. In the example, we show that informative data is often invisible for linearized methods that confuse zero correlation with statistical independence. Hence, PreDIA will often lead to substantially better sampling designs. Finally, we extend our example to specifically highlight the consideration of conceptual model uncertainty.

Maximum likelihood Bayesian model averaging and its predictive analysis for groundwater reactive transport models

USGS Publications Warehouse

Curtis, Gary P.; Lu, Dan; Ye, Ming

2015-01-01

While Bayesian model averaging (BMA) has been widely used in groundwater modeling, it is infrequently applied to groundwater reactive transport modeling because of multiple sources of uncertainty in the coupled hydrogeochemical processes and because of the long execution time of each model run. To resolve these problems, this study analyzed different levels of uncertainty in a hierarchical way, and used the maximum likelihood version of BMA, i.e., MLBMA, to improve the computational efficiency. This study demonstrates the applicability of MLBMA to groundwater reactive transport modeling in a synthetic case in which twenty-seven reactive transport models were designed to predict the reactive transport of hexavalent uranium (U(VI)) based on observations at a former uranium mill site near Naturita, CO. These reactive transport models contain three uncertain model components, i.e., parameterization of hydraulic conductivity, configuration of model boundary, and surface complexation reactions that simulate U(VI) adsorption. These uncertain model components were aggregated into the alternative models by integrating a hierarchical structure into MLBMA. The modeling results of the individual models and MLBMA were analyzed to investigate their predictive performance. The predictive logscore results show that MLBMA generally outperforms the best model, suggesting that using MLBMA is a sound strategy to achieve more robust model predictions relative to a single model. MLBMA works best when the alternative models are structurally distinct and have diverse model predictions. When correlation in model structure exists, two strategies were used to improve predictive performance by retaining structurally distinct models or assigning smaller prior model probabilities to correlated models. Since the synthetic models were designed using data from the Naturita site, the results of this study are expected to provide guidance for real-world modeling. Limitations of applying MLBMA to the synthetic study and future real-world modeling are discussed.

Genomic prediction using an iterative conditional expectation algorithm for a fast BayesC-like model.

PubMed

Dong, Linsong; Wang, Zhiyong

2018-06-11

Genomic prediction is feasible for estimating genomic breeding values because of dense genome-wide markers and credible statistical methods, such as Genomic Best Linear Unbiased Prediction (GBLUP) and various Bayesian methods. Compared with GBLUP, Bayesian methods propose more flexible assumptions for the distributions of SNP effects. However, most Bayesian methods are performed based on Markov chain Monte Carlo (MCMC) algorithms, leading to computational efficiency challenges. Hence, some fast Bayesian approaches, such as fast BayesB (fBayesB), were proposed to speed up the calculation. This study proposed another fast Bayesian method termed fast BayesC (fBayesC). The prior distribution of fBayesC assumes that a SNP with probability γ has a non-zero effect which comes from a normal density with a common variance. The simulated data from QTLMAS XII workshop and actual data on large yellow croaker were used to compare the predictive results of fBayesB, fBayesC and (MCMC-based) BayesC. The results showed that when γ was set as a small value, such as 0.01 in the simulated data or 0.001 in the actual data, fBayesB and fBayesC yielded lower prediction accuracies (abilities) than BayesC. In the actual data, fBayesC could yield very similar predictive abilities as BayesC when γ ≥ 0.01. When γ = 0.01, fBayesB could also yield similar results as fBayesC and BayesC. However, fBayesB could not yield an explicit result when γ ≥ 0.1, but a similar situation was not observed for fBayesC. Moreover, the computational speed of fBayesC was significantly faster than that of BayesC, making fBayesC a promising method for genomic prediction.

Bayesian analysis of heterogeneous treatment effects for patient-centered outcomes research.

PubMed

Henderson, Nicholas C; Louis, Thomas A; Wang, Chenguang; Varadhan, Ravi

2016-01-01

Evaluation of heterogeneity of treatment effect (HTE) is an essential aspect of personalized medicine and patient-centered outcomes research. Our goal in this article is to promote the use of Bayesian methods for subgroup analysis and to lower the barriers to their implementation by describing the ways in which the companion software beanz can facilitate these types of analyses. To advance this goal, we describe several key Bayesian models for investigating HTE and outline the ways in which they are well-suited to address many of the commonly cited challenges in the study of HTE. Topics highlighted include shrinkage estimation, model choice, sensitivity analysis, and posterior predictive checking. A case study is presented in which we demonstrate the use of the methods discussed.

Predictive model of outcome of targeted nodal assessment in colorectal cancer.

PubMed

Nissan, Aviram; Protic, Mladjan; Bilchik, Anton; Eberhardt, John; Peoples, George E; Stojadinovic, Alexander

2010-02-01

Improvement in staging accuracy is the principal aim of targeted nodal assessment in colorectal carcinoma. Technical factors independently predictive of false negative (FN) sentinel lymph node (SLN) mapping should be identified to facilitate operative decision making. To define independent predictors of FN SLN mapping and to develop a predictive model that could support surgical decisions. Data was analyzed from 2 completed prospective clinical trials involving 278 patients with colorectal carcinoma undergoing SLN mapping. Clinical outcome of interest was FN SLN(s), defined as one(s) with no apparent tumor cells in the presence of non-SLN metastases. To assess the independent predictive effect of a covariate for a nominal response (FN SLN), a logistic regression model was constructed and parameters estimated using maximum likelihood. A probabilistic Bayesian model was also trained and cross validated using 10-fold train-and-test sets to predict FN SLN mapping. Area under the curve (AUC) from receiver operating characteristics curves of these predictions was calculated to determine the predictive value of the model. Number of SLNs (<3; P = 0.03) and tumor-replaced nodes (P < 0.01) independently predicted FN SLN. Cross validation of the model created with Bayesian Network Analysis effectively predicted FN SLN (area under the curve = 0.84-0.86). The positive and negative predictive values of the model are 83% and 97%, respectively. This study supports a minimum threshold of 3 nodes for targeted nodal assessment in colorectal cancer, and establishes sufficient basis to conclude that SLN mapping and biopsy cannot be justified in the presence of clinically apparent tumor-replaced nodes.

Development of an internet based system for modeling biotin metabolism using Bayesian networks.

PubMed

Zhou, Jinglei; Wang, Dong; Schlegel, Vicki; Zempleni, Janos

2011-11-01

Biotin is an essential water-soluble vitamin crucial for maintaining normal body functions. The importance of biotin for human health has been under-appreciated but there is plenty of opportunity for future research with great importance for human health. Currently, carrying out predictions of biotin metabolism involves tedious manual manipulations. In this paper, we report the development of BiotinNet, an internet based program that uses Bayesian networks to integrate published data on various aspects of biotin metabolism. Users can provide a combination of values on the levels of biotin related metabolites to obtain the predictions on other metabolites that are not specified. As an inherent feature of Bayesian networks, the uncertainty of the prediction is also quantified and reported to the user. This program enables convenient in silico experiments regarding biotin metabolism, which can help researchers design future experiments while new data can be continuously incorporated. Copyright © 2011 Elsevier Ireland Ltd. All rights reserved.

Multi-dimensional classification of GABAergic interneurons with Bayesian network-modeled label uncertainty.

PubMed

Mihaljević, Bojan; Bielza, Concha; Benavides-Piccione, Ruth; DeFelipe, Javier; Larrañaga, Pedro

2014-01-01

Interneuron classification is an important and long-debated topic in neuroscience. A recent study provided a data set of digitally reconstructed interneurons classified by 42 leading neuroscientists according to a pragmatic classification scheme composed of five categorical variables, namely, of the interneuron type and four features of axonal morphology. From this data set we now learned a model which can classify interneurons, on the basis of their axonal morphometric parameters, into these five descriptive variables simultaneously. Because of differences in opinion among the neuroscientists, especially regarding neuronal type, for many interneurons we lacked a unique, agreed-upon classification, which we could use to guide model learning. Instead, we guided model learning with a probability distribution over the neuronal type and the axonal features, obtained, for each interneuron, from the neuroscientists' classification choices. We conveniently encoded such probability distributions with Bayesian networks, calling them label Bayesian networks (LBNs), and developed a method to predict them. This method predicts an LBN by forming a probabilistic consensus among the LBNs of the interneurons most similar to the one being classified. We used 18 axonal morphometric parameters as predictor variables, 13 of which we introduce in this paper as quantitative counterparts to the categorical axonal features. We were able to accurately predict interneuronal LBNs. Furthermore, when extracting crisp (i.e., non-probabilistic) predictions from the predicted LBNs, our method outperformed related work on interneuron classification. Our results indicate that our method is adequate for multi-dimensional classification of interneurons with probabilistic labels. Moreover, the introduced morphometric parameters are good predictors of interneuron type and the four features of axonal morphology and thus may serve as objective counterparts to the subjective, categorical axonal features.

Bayesian model aggregation for ensemble-based estimates of protein pKa values

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gosink, Luke J.; Hogan, Emilie A.; Pulsipher, Trenton C.

2014-03-01

This paper investigates an ensemble-based technique called Bayesian Model Averaging (BMA) to improve the performance of protein amino acid pmore » $$K_a$$ predictions. Structure-based p$$K_a$$ calculations play an important role in the mechanistic interpretation of protein structure and are also used to determine a wide range of protein properties. A diverse set of methods currently exist for p$$K_a$$ prediction, ranging from empirical statistical models to {\\it ab initio} quantum mechanical approaches. However, each of these methods are based on a set of assumptions that have inherent bias and sensitivities that can effect a model's accuracy and generalizability for p$$K_a$$ prediction in complicated biomolecular systems. We use BMA to combine eleven diverse prediction methods that each estimate pKa values of amino acids in staphylococcal nuclease. These methods are based on work conducted for the pKa Cooperative and the pKa measurements are based on experimental work conducted by the Garc{\\'i}a-Moreno lab. Our study demonstrates that the aggregated estimate obtained from BMA outperforms all individual prediction methods in our cross-validation study with improvements from 40-70\\% over other method classes. This work illustrates a new possible mechanism for improving the accuracy of p$$K_a$$ prediction and lays the foundation for future work on aggregate models that balance computational cost with prediction accuracy.« less

Bayesian component separation: The Planck experience

NASA Astrophysics Data System (ADS)

Wehus, Ingunn Kathrine; Eriksen, Hans Kristian

2018-05-01

Bayesian component separation techniques have played a central role in the data reduction process of Planck. The most important strength of this approach is its global nature, in which a parametric and physical model is fitted to the data. Such physical modeling allows the user to constrain very general data models, and jointly probe cosmological, astrophysical and instrumental parameters. This approach also supports statistically robust goodness-of-fit tests in terms of data-minus-model residual maps, which are essential for identifying residual systematic effects in the data. The main challenges are high code complexity and computational cost. Whether or not these costs are justified for a given experiment depends on its final uncertainty budget. We therefore predict that the importance of Bayesian component separation techniques is likely to increase with time for intensity mapping experiments, similar to what has happened in the CMB field, as observational techniques mature, and their overall sensitivity improves.

Bayesian calibration for forensic age estimation.

PubMed

Ferrante, Luigi; Skrami, Edlira; Gesuita, Rosaria; Cameriere, Roberto

2015-05-10

Forensic medicine is increasingly called upon to assess the age of individuals. Forensic age estimation is mostly required in relation to illegal immigration and identification of bodies or skeletal remains. A variety of age estimation methods are based on dental samples and use of regression models, where the age of an individual is predicted by morphological tooth changes that take place over time. From the medico-legal point of view, regression models, with age as the dependent random variable entail that age tends to be overestimated in the young and underestimated in the old. To overcome this bias, we describe a new full Bayesian calibration method (asymmetric Laplace Bayesian calibration) for forensic age estimation that uses asymmetric Laplace distribution as the probability model. The method was compared with three existing approaches (two Bayesian and a classical method) using simulated data. Although its accuracy was comparable with that of the other methods, the asymmetric Laplace Bayesian calibration appears to be significantly more reliable and robust in case of misspecification of the probability model. The proposed method was also applied to a real dataset of values of the pulp chamber of the right lower premolar measured on x-ray scans of individuals of known age. Copyright © 2015 John Wiley & Sons, Ltd.

Predicting nonstationary flood frequencies: Evidence supports an updated stationarity thesis in the United States

NASA Astrophysics Data System (ADS)

Luke, Adam; Vrugt, Jasper A.; AghaKouchak, Amir; Matthew, Richard; Sanders, Brett F.

2017-07-01

Nonstationary extreme value analysis (NEVA) can improve the statistical representation of observed flood peak distributions compared to stationary (ST) analysis, but management of flood risk relies on predictions of out-of-sample distributions for which NEVA has not been comprehensively evaluated. In this study, we apply split-sample testing to 1250 annual maximum discharge records in the United States and compare the predictive capabilities of NEVA relative to ST extreme value analysis using a log-Pearson Type III (LPIII) distribution. The parameters of the LPIII distribution in the ST and nonstationary (NS) models are estimated from the first half of each record using Bayesian inference. The second half of each record is reserved to evaluate the predictions under the ST and NS models. The NS model is applied for prediction by (1) extrapolating the trend of the NS model parameters throughout the evaluation period and (2) using the NS model parameter values at the end of the fitting period to predict with an updated ST model (uST). Our analysis shows that the ST predictions are preferred, overall. NS model parameter extrapolation is rarely preferred. However, if fitting period discharges are influenced by physical changes in the watershed, for example from anthropogenic activity, the uST model is strongly preferred relative to ST and NS predictions. The uST model is therefore recommended for evaluation of current flood risk in watersheds that have undergone physical changes. Supporting information includes a MATLAB® program that estimates the (ST/NS/uST) LPIII parameters from annual peak discharge data through Bayesian inference.

Impact of relationships between test and training animals and among training animals on reliability of genomic prediction.

PubMed

Wu, X; Lund, M S; Sun, D; Zhang, Q; Su, G

2015-10-01

One of the factors affecting the reliability of genomic prediction is the relationship among the animals of interest. This study investigated the reliability of genomic prediction in various scenarios with regard to the relationship between test and training animals, and among animals within the training data set. Different training data sets were generated from EuroGenomics data and a group of Nordic Holstein bulls (born in 2005 and afterwards) as a common test data set. Genomic breeding values were predicted using a genomic best linear unbiased prediction model and a Bayesian mixture model. The results showed that a closer relationship between test and training animals led to a higher reliability of genomic predictions for the test animals, while a closer relationship among training animals resulted in a lower reliability. In addition, the Bayesian mixture model in general led to a slightly higher reliability of genomic prediction, especially for the scenario of distant relationships between training and test animals. Therefore, to prevent a decrease in reliability, constant updates of the training population with animals from more recent generations are required. Moreover, a training population consisting of less-related animals is favourable for reliability of genomic prediction. © 2015 Blackwell Verlag GmbH.

Parallel Markov chain Monte Carlo - bridging the gap to high-performance Bayesian computation in animal breeding and genetics.

PubMed

Wu, Xiao-Lin; Sun, Chuanyu; Beissinger, Timothy M; Rosa, Guilherme Jm; Weigel, Kent A; Gatti, Natalia de Leon; Gianola, Daniel

2012-09-25

Most Bayesian models for the analysis of complex traits are not analytically tractable and inferences are based on computationally intensive techniques. This is true of Bayesian models for genome-enabled selection, which uses whole-genome molecular data to predict the genetic merit of candidate animals for breeding purposes. In this regard, parallel computing can overcome the bottlenecks that can arise from series computing. Hence, a major goal of the present study is to bridge the gap to high-performance Bayesian computation in the context of animal breeding and genetics. Parallel Monte Carlo Markov chain algorithms and strategies are described in the context of animal breeding and genetics. Parallel Monte Carlo algorithms are introduced as a starting point including their applications to computing single-parameter and certain multiple-parameter models. Then, two basic approaches for parallel Markov chain Monte Carlo are described: one aims at parallelization within a single chain; the other is based on running multiple chains, yet some variants are discussed as well. Features and strategies of the parallel Markov chain Monte Carlo are illustrated using real data, including a large beef cattle dataset with 50K SNP genotypes. Parallel Markov chain Monte Carlo algorithms are useful for computing complex Bayesian models, which does not only lead to a dramatic speedup in computing but can also be used to optimize model parameters in complex Bayesian models. Hence, we anticipate that use of parallel Markov chain Monte Carlo will have a profound impact on revolutionizing the computational tools for genomic selection programs.

Parallel Markov chain Monte Carlo - bridging the gap to high-performance Bayesian computation in animal breeding and genetics

PubMed Central

2012-01-01

Background Most Bayesian models for the analysis of complex traits are not analytically tractable and inferences are based on computationally intensive techniques. This is true of Bayesian models for genome-enabled selection, which uses whole-genome molecular data to predict the genetic merit of candidate animals for breeding purposes. In this regard, parallel computing can overcome the bottlenecks that can arise from series computing. Hence, a major goal of the present study is to bridge the gap to high-performance Bayesian computation in the context of animal breeding and genetics. Results Parallel Monte Carlo Markov chain algorithms and strategies are described in the context of animal breeding and genetics. Parallel Monte Carlo algorithms are introduced as a starting point including their applications to computing single-parameter and certain multiple-parameter models. Then, two basic approaches for parallel Markov chain Monte Carlo are described: one aims at parallelization within a single chain; the other is based on running multiple chains, yet some variants are discussed as well. Features and strategies of the parallel Markov chain Monte Carlo are illustrated using real data, including a large beef cattle dataset with 50K SNP genotypes. Conclusions Parallel Markov chain Monte Carlo algorithms are useful for computing complex Bayesian models, which does not only lead to a dramatic speedup in computing but can also be used to optimize model parameters in complex Bayesian models. Hence, we anticipate that use of parallel Markov chain Monte Carlo will have a profound impact on revolutionizing the computational tools for genomic selection programs. PMID:23009363

Bayesian Maximum Entropy Integration of Ozone Observations and Model Predictions: A National Application.

PubMed

Xu, Yadong; Serre, Marc L; Reyes, Jeanette; Vizuete, William

2016-04-19

To improve ozone exposure estimates for ambient concentrations at a national scale, we introduce our novel Regionalized Air Quality Model Performance (RAMP) approach to integrate chemical transport model (CTM) predictions with the available ozone observations using the Bayesian Maximum Entropy (BME) framework. The framework models the nonlinear and nonhomoscedastic relation between air pollution observations and CTM predictions and for the first time accounts for variability in CTM model performance. A validation analysis using only noncollocated data outside of a validation radius rv was performed and the R(2) between observations and re-estimated values for two daily metrics, the daily maximum 8-h average (DM8A) and the daily 24-h average (D24A) ozone concentrations, were obtained with the OBS scenario using ozone observations only in contrast with the RAMP and a Constant Air Quality Model Performance (CAMP) scenarios. We show that, by accounting for the spatial and temporal variability in model performance, our novel RAMP approach is able to extract more information in terms of R(2) increase percentage, with over 12 times for the DM8A and over 3.5 times for the D24A ozone concentrations, from CTM predictions than the CAMP approach assuming that model performance does not change across space and time.

Building Models to Predict Hint-or-Attempt Actions of Students

ERIC Educational Resources Information Center

Castro, Francisco Enrique Vicente; Adjei, Seth; Colombo, Tyler; Heffernan, Neil

2015-01-01

A great deal of research in educational data mining is geared towards predicting student performance. Bayesian Knowledge Tracing, Performance Factors Analysis, and the different variations of these have been introduced and have had some success at predicting student knowledge. It is worth noting, however, that very little has been done to…

Bayesian Monte Carlo and Maximum Likelihood Approach for ...

EPA Pesticide Factsheets

Model uncertainty estimation and risk assessment is essential to environmental management and informed decision making on pollution mitigation strategies. In this study, we apply a probabilistic methodology, which combines Bayesian Monte Carlo simulation and Maximum Likelihood estimation (BMCML) to calibrate a lake oxygen recovery model. We first derive an analytical solution of the differential equation governing lake-averaged oxygen dynamics as a function of time-variable wind speed. Statistical inferences on model parameters and predictive uncertainty are then drawn by Bayesian conditioning of the analytical solution on observed daily wind speed and oxygen concentration data obtained from an earlier study during two recovery periods on a eutrophic lake in upper state New York. The model is calibrated using oxygen recovery data for one year and statistical inferences were validated using recovery data for another year. Compared with essentially two-step, regression and optimization approach, the BMCML results are more comprehensive and performed relatively better in predicting the observed temporal dissolved oxygen levels (DO) in the lake. BMCML also produced comparable calibration and validation results with those obtained using popular Markov Chain Monte Carlo technique (MCMC) and is computationally simpler and easier to implement than the MCMC. Next, using the calibrated model, we derive an optimal relationship between liquid film-transfer coefficien

Bayesian outcome-based strategy classification.

PubMed

Lee, Michael D

2016-03-01

Hilbig and Moshagen (Psychonomic Bulletin & Review, 21, 1431-1443, 2014) recently developed a method for making inferences about the decision processes people use in multi-attribute forced choice tasks. Their paper makes a number of worthwhile theoretical and methodological contributions. Theoretically, they provide an insightful psychological motivation for a probabilistic extension of the widely-used "weighted additive" (WADD) model, and show how this model, as well as other important models like "take-the-best" (TTB), can and should be expressed in terms of meaningful priors. Methodologically, they develop an inference approach based on the Minimum Description Length (MDL) principles that balances both the goodness-of-fit and complexity of the decision models they consider. This paper aims to preserve these useful contributions, but provide a complementary Bayesian approach with some theoretical and methodological advantages. We develop a simple graphical model, implemented in JAGS, that allows for fully Bayesian inferences about which models people use to make decisions. To demonstrate the Bayesian approach, we apply it to the models and data considered by Hilbig and Moshagen (Psychonomic Bulletin & Review, 21, 1431-1443, 2014), showing how a prior predictive analysis of the models, and posterior inferences about which models people use and the parameter settings at which they use them, can contribute to our understanding of human decision making.

Performance assessment of a Bayesian Forecasting System (BFS) for real-time flood forecasting

NASA Astrophysics Data System (ADS)

Biondi, D.; De Luca, D. L.

2013-02-01

SummaryThe paper evaluates, for a number of flood events, the performance of a Bayesian Forecasting System (BFS), with the aim of evaluating total uncertainty in real-time flood forecasting. The predictive uncertainty of future streamflow is estimated through the Bayesian integration of two separate processors. The former evaluates the propagation of input uncertainty on simulated river discharge, the latter computes the hydrological uncertainty of actual river discharge associated with all other possible sources of error. A stochastic model and a distributed rainfall-runoff model were assumed, respectively, for rainfall and hydrological response simulations. A case study was carried out for a small basin in the Calabria region (southern Italy). The performance assessment of the BFS was performed with adequate verification tools suited for probabilistic forecasts of continuous variables such as streamflow. Graphical tools and scalar metrics were used to evaluate several attributes of the forecast quality of the entire time-varying predictive distributions: calibration, sharpness, accuracy, and continuous ranked probability score (CRPS). Besides the overall system, which incorporates both sources of uncertainty, other hypotheses resulting from the BFS properties were examined, corresponding to (i) a perfect hydrological model; (ii) a non-informative rainfall forecast for predicting streamflow; and (iii) a perfect input forecast. The results emphasize the importance of using different diagnostic approaches to perform comprehensive analyses of predictive distributions, to arrive at a multifaceted view of the attributes of the prediction. For the case study, the selected criteria revealed the interaction of the different sources of error, in particular the crucial role of the hydrological uncertainty processor when compensating, at the cost of wider forecast intervals, for the unreliable and biased predictive distribution resulting from the Precipitation Uncertainty Processor.

A Bayesian network approach to predicting nest presence of thefederally-threatened piping plover (Charadrius melodus) using barrier island features

USGS Publications Warehouse

Gieder, Katherina D.; Karpanty, Sarah M.; Fraser, James D.; Catlin, Daniel H.; Gutierrez, Benjamin T.; Plant, Nathaniel G.; Turecek, Aaron M.; Thieler, E. Robert

2014-01-01

Sea-level rise and human development pose significant threats to shorebirds, particularly for species that utilize barrier island habitat. The piping plover (Charadrius melodus) is a federally-listed shorebird that nests on barrier islands and rapidly responds to changes in its physical environment, making it an excellent species with which to model how shorebird species may respond to habitat change related to sea-level rise and human development. The uncertainty and complexity in predicting sea-level rise, the responses of barrier island habitats to sea-level rise, and the responses of species to sea-level rise and human development necessitate a modelling approach that can link species to the physical habitat features that will be altered by changes in sea level and human development. We used a Bayesian network framework to develop a model that links piping plover nest presence to the physical features of their nesting habitat on a barrier island that is impacted by sea-level rise and human development, using three years of data (1999, 2002, and 2008) from Assateague Island National Seashore in Maryland. Our model performance results showed that we were able to successfully predict nest presence given a wide range of physical conditions within the model’s dataset. We found that model predictions were more successful when the range of physical conditions included in model development was varied rather than when those physical conditions were narrow. We also found that all model predictions had fewer false negatives (nests predicted to be absent when they were actually present in the dataset) than false positives (nests predicted to be present when they were actually absent in the dataset), indicating that our model correctly predicted nest presence better than nest absence. These results indicated that our approach of using a Bayesian network to link specific physical features to nest presence will be useful for modelling impacts of sea-level rise- or human-related habitat change on barrier islands. We recommend that potential users of this method utilize multiple years of data that represent a wide range of physical conditions in model development, because the model performed less well when constructed using a narrow range of physical conditions. Further, given that there will always be some uncertainty in predictions of future physical habitat conditions related to sea-level rise and/or human development, predictive models will perform best when developed using multiple, varied years of data input.

Development of a clinical decision model for thyroid nodules

PubMed Central

Stojadinovic, Alexander; Peoples, George E; Libutti, Steven K; Henry, Leonard R; Eberhardt, John; Howard, Robin S; Gur, David; Elster, Eric A; Nissan, Aviram

2009-01-01

Background Thyroid nodules represent a common problem brought to medical attention. Four to seven percent of the United States adult population (10–18 million people) has a palpable thyroid nodule, however the majority (>95%) of thyroid nodules are benign. While, fine needle aspiration remains the most cost effective and accurate diagnostic tool for thyroid nodules in current practice, over 20% of patients undergoing FNA of a thyroid nodule have indeterminate cytology (follicular neoplasm) with associated malignancy risk prevalence of 20–30%. These patients require thyroid lobectomy/isthmusectomy purely for the purpose of attaining a definitive diagnosis. Given that the majority (70–80%) of these patients have benign surgical pathology, thyroidectomy in these patients is conducted principally with diagnostic intent. Clinical models predictive of malignancy risk are needed to support treatment decisions in patients with thyroid nodules in order to reduce morbidity associated with unnecessary diagnostic surgery. Methods Data were analyzed from a completed prospective cohort trial conducted over a 4-year period involving 216 patients with thyroid nodules undergoing ultrasound (US), electrical impedance scanning (EIS) and fine needle aspiration cytology (FNA) prior to thyroidectomy. A Bayesian model was designed to predict malignancy in thyroid nodules based on multivariate dependence relationships between independent covariates. Ten-fold cross-validation was performed to estimate classifier error wherein the data set was randomized into ten separate and unique train and test sets consisting of a training set (90% of records) and a test set (10% of records). A receiver-operating-characteristics (ROC) curve of these predictions and area under the curve (AUC) were calculated to determine model robustness for predicting malignancy in thyroid nodules. Results Thyroid nodule size, FNA cytology, US and EIS characteristics were highly predictive of malignancy. Cross validation of the model created with Bayesian Network Analysis effectively predicted malignancy [AUC = 0.88 (95%CI: 0.82–0.94)] in thyroid nodules. The positive and negative predictive values of the model are 83% (95%CI: 76%–91%) and 79% (95%CI: 72%–86%), respectively. Conclusion An integrated predictive decision model using Bayesian inference incorporating readily obtainable thyroid nodule measures is clinically relevant, as it effectively predicts malignancy in thyroid nodules. This model warrants further validation testing in prospective clinical trials. PMID:19664278

Bayesian state space models for dynamic genetic network construction across multiple tissues.

PubMed

Liang, Yulan; Kelemen, Arpad

2016-08-01

Construction of gene-gene interaction networks and potential pathways is a challenging and important problem in genomic research for complex diseases while estimating the dynamic changes of the temporal correlations and non-stationarity are the keys in this process. In this paper, we develop dynamic state space models with hierarchical Bayesian settings to tackle this challenge for inferring the dynamic profiles and genetic networks associated with disease treatments. We treat both the stochastic transition matrix and the observation matrix time-variant and include temporal correlation structures in the covariance matrix estimations in the multivariate Bayesian state space models. The unevenly spaced short time courses with unseen time points are treated as hidden state variables. Hierarchical Bayesian approaches with various prior and hyper-prior models with Monte Carlo Markov Chain and Gibbs sampling algorithms are used to estimate the model parameters and the hidden state variables. We apply the proposed Hierarchical Bayesian state space models to multiple tissues (liver, skeletal muscle, and kidney) Affymetrix time course data sets following corticosteroid (CS) drug administration. Both simulation and real data analysis results show that the genomic changes over time and gene-gene interaction in response to CS treatment can be well captured by the proposed models. The proposed dynamic Hierarchical Bayesian state space modeling approaches could be expanded and applied to other large scale genomic data, such as next generation sequence (NGS) combined with real time and time varying electronic health record (EHR) for more comprehensive and robust systematic and network based analysis in order to transform big biomedical data into predictions and diagnostics for precision medicine and personalized healthcare with better decision making and patient outcomes.

Optimization of global model composed of radial basis functions using the term-ranking approach

DOE Office of Scientific and Technical Information (OSTI.GOV)

Cai, Peng; Tao, Chao, E-mail: taochao@nju.edu.cn; Liu, Xiao-Jun

2014-03-15

A term-ranking method is put forward to optimize the global model composed of radial basis functions to improve the predictability of the model. The effectiveness of the proposed method is examined by numerical simulation and experimental data. Numerical simulations indicate that this method can significantly lengthen the prediction time and decrease the Bayesian information criterion of the model. The application to real voice signal shows that the optimized global model can capture more predictable component in chaos-like voice data and simultaneously reduce the predictable component (periodic pitch) in the residual signal.

A Bayesian network model for predicting type 2 diabetes risk based on electronic health records

NASA Astrophysics Data System (ADS)

Xie, Jiang; Liu, Yan; Zeng, Xu; Zhang, Wu; Mei, Zhen

2017-07-01

An extensive, in-depth study of diabetes risk factors (DBRF) is of crucial importance to prevent (or reduce) the chance of suffering from type 2 diabetes (T2D). Accumulation of electronic health records (EHRs) makes it possible to build nonlinear relationships between risk factors and diabetes. However, the current DBRF researches mainly focus on qualitative analyses, and the inconformity of physical examination items makes the risk factors likely to be lost, which drives us to study the novel machine learning approach for risk model development. In this paper, we use Bayesian networks (BNs) to analyze the relationship between physical examination information and T2D, and to quantify the link between risk factors and T2D. Furthermore, with the quantitative analyses of DBRF, we adopt EHR and propose a machine learning approach based on BNs to predict the risk of T2D. The experiments demonstrate that our approach can lead to better predictive performance than the classical risk model.

Drug delivery optimization through Bayesian networks.

PubMed Central

Bellazzi, R.

1992-01-01

This paper describes how Bayesian Networks can be used in combination with compartmental models to plan Recombinant Human Erythropoietin (r-HuEPO) delivery in the treatment of anemia of chronic uremic patients. Past measurements of hematocrit or hemoglobin concentration in a patient during the therapy can be exploited to adjust the parameters of a compartmental model of the erythropoiesis. This adaptive process allows more accurate patient-specific predictions, and hence a more rational dosage planning. We describe a drug delivery optimization protocol, based on our approach. Some results obtained on real data are presented. PMID:1482938

Bayesian inference of physiologically meaningful parameters from body sway measurements.

PubMed

Tietäväinen, A; Gutmann, M U; Keski-Vakkuri, E; Corander, J; Hæggström, E

2017-06-19

The control of the human body sway by the central nervous system, muscles, and conscious brain is of interest since body sway carries information about the physiological status of a person. Several models have been proposed to describe body sway in an upright standing position, however, due to the statistical intractability of the more realistic models, no formal parameter inference has previously been conducted and the expressive power of such models for real human subjects remains unknown. Using the latest advances in Bayesian statistical inference for intractable models, we fitted a nonlinear control model to posturographic measurements, and we showed that it can accurately predict the sway characteristics of both simulated and real subjects. Our method provides a full statistical characterization of the uncertainty related to all model parameters as quantified by posterior probability density functions, which is useful for comparisons across subjects and test settings. The ability to infer intractable control models from sensor data opens new possibilities for monitoring and predicting body status in health applications.

A Bayesian Poisson-lognormal Model for Count Data for Multiple-Trait Multiple-Environment Genomic-Enabled Prediction

PubMed Central

Montesinos-López, Osval A.; Montesinos-López, Abelardo; Crossa, José; Toledo, Fernando H.; Montesinos-López, José C.; Singh, Pawan; Juliana, Philomin; Salinas-Ruiz, Josafhat

2017-01-01

When a plant scientist wishes to make genomic-enabled predictions of multiple traits measured in multiple individuals in multiple environments, the most common strategy for performing the analysis is to use a single trait at a time taking into account genotype × environment interaction (G × E), because there is a lack of comprehensive models that simultaneously take into account the correlated counting traits and G × E. For this reason, in this study we propose a multiple-trait and multiple-environment model for count data. The proposed model was developed under the Bayesian paradigm for which we developed a Markov Chain Monte Carlo (MCMC) with noninformative priors. This allows obtaining all required full conditional distributions of the parameters leading to an exact Gibbs sampler for the posterior distribution. Our model was tested with simulated data and a real data set. Results show that the proposed multi-trait, multi-environment model is an attractive alternative for modeling multiple count traits measured in multiple environments. PMID:28364037

Evaluating Bayesian spatial methods for modelling species distributions with clumped and restricted occurrence data.

PubMed

Redding, David W; Lucas, Tim C D; Blackburn, Tim M; Jones, Kate E

2017-01-01

Statistical approaches for inferring the spatial distribution of taxa (Species Distribution Models, SDMs) commonly rely on available occurrence data, which is often clumped and geographically restricted. Although available SDM methods address some of these factors, they could be more directly and accurately modelled using a spatially-explicit approach. Software to fit models with spatial autocorrelation parameters in SDMs are now widely available, but whether such approaches for inferring SDMs aid predictions compared to other methodologies is unknown. Here, within a simulated environment using 1000 generated species' ranges, we compared the performance of two commonly used non-spatial SDM methods (Maximum Entropy Modelling, MAXENT and boosted regression trees, BRT), to a spatial Bayesian SDM method (fitted using R-INLA), when the underlying data exhibit varying combinations of clumping and geographic restriction. Finally, we tested how any recommended methodological settings designed to account for spatially non-random patterns in the data impact inference. Spatial Bayesian SDM method was the most consistently accurate method, being in the top 2 most accurate methods in 7 out of 8 data sampling scenarios. Within high-coverage sample datasets, all methods performed fairly similarly. When sampling points were randomly spread, BRT had a 1-3% greater accuracy over the other methods and when samples were clumped, the spatial Bayesian SDM method had a 4%-8% better AUC score. Alternatively, when sampling points were restricted to a small section of the true range all methods were on average 10-12% less accurate, with greater variation among the methods. Model inference under the recommended settings to account for autocorrelation was not impacted by clumping or restriction of data, except for the complexity of the spatial regression term in the spatial Bayesian model. Methods, such as those made available by R-INLA, can be successfully used to account for spatial autocorrelation in an SDM context and, by taking account of random effects, produce outputs that can better elucidate the role of covariates in predicting species occurrence. Given that it is often unclear what the drivers are behind data clumping in an empirical occurrence dataset, or indeed how geographically restricted these data are, spatially-explicit Bayesian SDMs may be the better choice when modelling the spatial distribution of target species.

Multivariable and Bayesian Network Analysis of Outcome Predictors in Acute Aneurysmal Subarachnoid Hemorrhage: Review of a Pure Surgical Series in the Post-International Subarachnoid Aneurysm Trial Era.

PubMed

Zador, Zsolt; Huang, Wendy; Sperrin, Matthew; Lawton, Michael T

2018-06-01

Following the International Subarachnoid Aneurysm Trial (ISAT), evolving treatment modalities for acute aneurysmal subarachnoid hemorrhage (aSAH) has changed the case mix of patients undergoing urgent surgical clipping. To update our knowledge on outcome predictors by analyzing admission parameters in a pure surgical series using variable importance ranking and machine learning. We reviewed a single surgeon's case series of 226 patients suffering from aSAH treated with urgent surgical clipping. Predictions were made using logistic regression models, and predictive performance was assessed using areas under the receiver operating curve (AUC). We established variable importance ranking using partial Nagelkerke R2 scores. Probabilistic associations between variables were depicted using Bayesian networks, a method of machine learning. Importance ranking showed that World Federation of Neurosurgical Societies (WFNS) grade and age were the most influential outcome prognosticators. Inclusion of only these 2 predictors was sufficient to maintain model performance compared to when all variables were considered (AUC = 0.8222, 95% confidence interval (CI): 0.7646-0.88 vs 0.8218, 95% CI: 0.7616-0.8821, respectively, DeLong's P = .992). Bayesian networks showed that age and WFNS grade were associated with several variables such as laboratory results and cardiorespiratory parameters. Our study is the first to report early outcomes and formal predictor importance ranking following aSAH in a post-ISAT surgical case series. Models showed good predictive power with fewer relevant predictors than in similar size series. Bayesian networks proved to be a powerful tool in visualizing the widespread association of the 2 key predictors with admission variables, explaining their importance and demonstrating the potential for hypothesis generation.

Cortical Hierarchies Perform Bayesian Causal Inference in Multisensory Perception

PubMed Central

Rohe, Tim; Noppeney, Uta

2015-01-01

To form a veridical percept of the environment, the brain needs to integrate sensory signals from a common source but segregate those from independent sources. Thus, perception inherently relies on solving the “causal inference problem.” Behaviorally, humans solve this problem optimally as predicted by Bayesian Causal Inference; yet, the underlying neural mechanisms are unexplored. Combining psychophysics, Bayesian modeling, functional magnetic resonance imaging (fMRI), and multivariate decoding in an audiovisual spatial localization task, we demonstrate that Bayesian Causal Inference is performed by a hierarchy of multisensory processes in the human brain. At the bottom of the hierarchy, in auditory and visual areas, location is represented on the basis that the two signals are generated by independent sources (= segregation). At the next stage, in posterior intraparietal sulcus, location is estimated under the assumption that the two signals are from a common source (= forced fusion). Only at the top of the hierarchy, in anterior intraparietal sulcus, the uncertainty about the causal structure of the world is taken into account and sensory signals are combined as predicted by Bayesian Causal Inference. Characterizing the computational operations of signal interactions reveals the hierarchical nature of multisensory perception in human neocortex. It unravels how the brain accomplishes Bayesian Causal Inference, a statistical computation fundamental for perception and cognition. Our results demonstrate how the brain combines information in the face of uncertainty about the underlying causal structure of the world. PMID:25710328

Spatiotemporal modeling of PM2.5 concentrations at the national scale combining land use regression and Bayesian maximum entropy in China.

PubMed

Chen, Li; Gao, Shuang; Zhang, Hui; Sun, Yanling; Ma, Zhenxing; Vedal, Sverre; Mao, Jian; Bai, Zhipeng

2018-05-03

Concentrations of particulate matter with aerodynamic diameter <2.5 μm (PM 2.5 ) are relatively high in China. Estimation of PM 2.5 exposure is complex because PM 2.5 exhibits complex spatiotemporal patterns. To improve the validity of exposure predictions, several methods have been developed and applied worldwide. A hybrid approach combining a land use regression (LUR) model and Bayesian Maximum Entropy (BME) interpolation of the LUR space-time residuals were developed to estimate the PM 2.5 concentrations on a national scale in China. This hybrid model could potentially provide more valid predictions than a commonly-used LUR model. The LUR/BME model had good performance characteristics, with R 2  = 0.82 and root mean square error (RMSE) of 4.6 μg/m 3 . Prediction errors of the LUR/BME model were reduced by incorporating soft data accounting for data uncertainty, with the R 2 increasing by 6%. The performance of LUR/BME is better than OK/BME. The LUR/BME model is the most accurate fine spatial scale PM 2.5 model developed to date for China. Copyright © 2018. Published by Elsevier Ltd.

Multi-objective calibration and uncertainty analysis of hydrologic models; A comparative study between formal and informal methods

NASA Astrophysics Data System (ADS)

Shafii, M.; Tolson, B.; Matott, L. S.

2012-04-01

Hydrologic modeling has benefited from significant developments over the past two decades. This has resulted in building of higher levels of complexity into hydrologic models, which eventually makes the model evaluation process (parameter estimation via calibration and uncertainty analysis) more challenging. In order to avoid unreasonable parameter estimates, many researchers have suggested implementation of multi-criteria calibration schemes. Furthermore, for predictive hydrologic models to be useful, proper consideration of uncertainty is essential. Consequently, recent research has emphasized comprehensive model assessment procedures in which multi-criteria parameter estimation is combined with statistically-based uncertainty analysis routines such as Bayesian inference using Markov Chain Monte Carlo (MCMC) sampling. Such a procedure relies on the use of formal likelihood functions based on statistical assumptions, and moreover, the Bayesian inference structured on MCMC samplers requires a considerably large number of simulations. Due to these issues, especially in complex non-linear hydrological models, a variety of alternative informal approaches have been proposed for uncertainty analysis in the multi-criteria context. This study aims at exploring a number of such informal uncertainty analysis techniques in multi-criteria calibration of hydrological models. The informal methods addressed in this study are (i) Pareto optimality which quantifies the parameter uncertainty using the Pareto solutions, (ii) DDS-AU which uses the weighted sum of objective functions to derive the prediction limits, and (iii) GLUE which describes the total uncertainty through identification of behavioral solutions. The main objective is to compare such methods with MCMC-based Bayesian inference with respect to factors such as computational burden, and predictive capacity, which are evaluated based on multiple comparative measures. The measures for comparison are calculated both for calibration and evaluation periods. The uncertainty analysis methodologies are applied to a simple 5-parameter rainfall-runoff model, called HYMOD.

Ridge, Lasso and Bayesian additive-dominance genomic models.

PubMed

Azevedo, Camila Ferreira; de Resende, Marcos Deon Vilela; E Silva, Fabyano Fonseca; Viana, José Marcelo Soriano; Valente, Magno Sávio Ferreira; Resende, Márcio Fernando Ribeiro; Muñoz, Patricio

2015-08-25

A complete approach for genome-wide selection (GWS) involves reliable statistical genetics models and methods. Reports on this topic are common for additive genetic models but not for additive-dominance models. The objective of this paper was (i) to compare the performance of 10 additive-dominance predictive models (including current models and proposed modifications), fitted using Bayesian, Lasso and Ridge regression approaches; and (ii) to decompose genomic heritability and accuracy in terms of three quantitative genetic information sources, namely, linkage disequilibrium (LD), co-segregation (CS) and pedigree relationships or family structure (PR). The simulation study considered two broad sense heritability levels (0.30 and 0.50, associated with narrow sense heritabilities of 0.20 and 0.35, respectively) and two genetic architectures for traits (the first consisting of small gene effects and the second consisting of a mixed inheritance model with five major genes). G-REML/G-BLUP and a modified Bayesian/Lasso (called BayesA*B* or t-BLASSO) method performed best in the prediction of genomic breeding as well as the total genotypic values of individuals in all four scenarios (two heritabilities x two genetic architectures). The BayesA*B*-type method showed a better ability to recover the dominance variance/additive variance ratio. Decomposition of genomic heritability and accuracy revealed the following descending importance order of information: LD, CS and PR not captured by markers, the last two being very close. Amongst the 10 models/methods evaluated, the G-BLUP, BAYESA*B* (-2,8) and BAYESA*B* (4,6) methods presented the best results and were found to be adequate for accurately predicting genomic breeding and total genotypic values as well as for estimating additive and dominance in additive-dominance genomic models.

Uncertainty analysis of depth predictions from seismic reflection data using Bayesian statistics

NASA Astrophysics Data System (ADS)

Michelioudakis, Dimitrios G.; Hobbs, Richard W.; Caiado, Camila C. S.

2018-03-01

Estimating the depths of target horizons from seismic reflection data is an important task in exploration geophysics. To constrain these depths we need a reliable and accurate velocity model. Here, we build an optimum 2D seismic reflection data processing flow focused on pre - stack deghosting filters and velocity model building and apply Bayesian methods, including Gaussian process emulation and Bayesian History Matching (BHM), to estimate the uncertainties of the depths of key horizons near the borehole DSDP-258 located in the Mentelle Basin, south west of Australia, and compare the results with the drilled core from that well. Following this strategy, the tie between the modelled and observed depths from DSDP-258 core was in accordance with the ± 2σ posterior credibility intervals and predictions for depths to key horizons were made for the two new drill sites, adjacent the existing borehole of the area. The probabilistic analysis allowed us to generate multiple realizations of pre-stack depth migrated images, these can be directly used to better constrain interpretation and identify potential risk at drill sites. The method will be applied to constrain the drilling targets for the upcoming International Ocean Discovery Program (IODP), leg 369.

Uncertainty analysis of depth predictions from seismic reflection data using Bayesian statistics

NASA Astrophysics Data System (ADS)

Michelioudakis, Dimitrios G.; Hobbs, Richard W.; Caiado, Camila C. S.

2018-06-01

Estimating the depths of target horizons from seismic reflection data is an important task in exploration geophysics. To constrain these depths we need a reliable and accurate velocity model. Here, we build an optimum 2-D seismic reflection data processing flow focused on pre-stack deghosting filters and velocity model building and apply Bayesian methods, including Gaussian process emulation and Bayesian History Matching, to estimate the uncertainties of the depths of key horizons near the Deep Sea Drilling Project (DSDP) borehole 258 (DSDP-258) located in the Mentelle Basin, southwest of Australia, and compare the results with the drilled core from that well. Following this strategy, the tie between the modelled and observed depths from DSDP-258 core was in accordance with the ±2σ posterior credibility intervals and predictions for depths to key horizons were made for the two new drill sites, adjacent to the existing borehole of the area. The probabilistic analysis allowed us to generate multiple realizations of pre-stack depth migrated images, these can be directly used to better constrain interpretation and identify potential risk at drill sites. The method will be applied to constrain the drilling targets for the upcoming International Ocean Discovery Program, leg 369.

Impact of Colic Pain as a Significant Factor for Predicting the Stone Free Rate of One-Session Shock Wave Lithotripsy for Treating Ureter Stones: A Bayesian Logistic Regression Model Analysis

PubMed Central

Chung, Doo Yong; Cho, Kang Su; Lee, Dae Hun; Han, Jang Hee; Kang, Dong Hyuk; Jung, Hae Do; Kown, Jong Kyou; Ham, Won Sik; Choi, Young Deuk; Lee, Joo Yong

2015-01-01

Purpose This study was conducted to evaluate colic pain as a prognostic pretreatment factor that can influence ureter stone clearance and to estimate the probability of stone-free status in shock wave lithotripsy (SWL) patients with a ureter stone. Materials and Methods We retrospectively reviewed the medical records of 1,418 patients who underwent their first SWL between 2005 and 2013. Among these patients, 551 had a ureter stone measuring 4–20 mm and were thus eligible for our analyses. The colic pain as the chief complaint was defined as either subjective flank pain during history taking and physical examination. Propensity-scores for established for colic pain was calculated for each patient using multivariate logistic regression based upon the following covariates: age, maximal stone length (MSL), and mean stone density (MSD). Each factor was evaluated as predictor for stone-free status by Bayesian and non-Bayesian logistic regression model. Results After propensity-score matching, 217 patients were extracted in each group from the total patient cohort. There were no statistical differences in variables used in propensity- score matching. One-session success and stone-free rate were also higher in the painful group (73.7% and 71.0%, respectively) than in the painless group (63.6% and 60.4%, respectively). In multivariate non-Bayesian and Bayesian logistic regression models, a painful stone, shorter MSL, and lower MSD were significant factors for one-session stone-free status in patients who underwent SWL. Conclusions Colic pain in patients with ureter calculi was one of the significant predicting factors including MSL and MSD for one-session stone-free status of SWL. PMID:25902059

A Bayesian Hierarchical Modeling Approach to Predicting Flow in Ungauged Basins

EPA Science Inventory

Recent innovative approaches to identifying and applying regression-based relationships between land use patterns (such as increasing impervious surface area and decreasing vegetative cover) and rainfall-runoff model parameters represent novel and promising improvements to predic...

Probabilistic estimation of dune retreat on the Gold Coast, Australia

USGS Publications Warehouse

Palmsten, Margaret L.; Splinter, Kristen D.; Plant, Nathaniel G.; Stockdon, Hilary F.

2014-01-01

Sand dunes are an important natural buffer between storm impacts and development backing the beach on the Gold Coast of Queensland, Australia. The ability to forecast dune erosion at a prediction horizon of days to a week would allow efficient and timely response to dune erosion in this highly populated area. Towards this goal, we modified an existing probabilistic dune erosion model for use on the Gold Coast. The original model was trained using observations of dune response from Hurricane Ivan on Santa Rosa Island, Florida, USA (Plant and Stockdon 2012. Probabilistic prediction of barrier-island response to hurricanes, Journal of Geophysical Research, 117(F3), F03015). The model relates dune position change to pre-storm dune elevations, dune widths, and beach widths, along with storm surge and run-up using a Bayesian network. The Bayesian approach captures the uncertainty of inputs and predictions through the conditional probabilities between variables. Three versions of the barrier island response Bayesian network were tested for use on the Gold Coast. One network has the same structure as the original and was trained with the Santa Rosa Island data. The second network has a modified design and was trained using only pre- and post-storm data from 1988-2009 for the Gold Coast. The third version of the network has the same design as the second version of the network and was trained with the combined data from the Gold Coast and Santa Rosa Island. The two networks modified for use on the Gold Coast hindcast dune retreat with equal accuracy. Both networks explained 60% of the observed dune retreat variance, which is comparable to the skill observed by Plant and Stockdon (2012) in the initial Bayesian network application at Santa Rosa Island. The new networks improved predictions relative to application of the original network on the Gold Coast. Dune width was the most important morphologic variable in hindcasting dune retreat, while hydrodynamic variables, surge and run-up elevation, were also important

Spatial mapping and prediction of Plasmodium falciparum infection risk among school-aged children in Côte d'Ivoire.

PubMed

Houngbedji, Clarisse A; Chammartin, Frédérique; Yapi, Richard B; Hürlimann, Eveline; N'Dri, Prisca B; Silué, Kigbafori D; Soro, Gotianwa; Koudou, Benjamin G; Assi, Serge-Brice; N'Goran, Eliézer K; Fantodji, Agathe; Utzinger, Jürg; Vounatsou, Penelope; Raso, Giovanna

2016-09-07

In Côte d'Ivoire, malaria remains a major public health issue, and thus a priority to be tackled. The aim of this study was to identify spatially explicit indicators of Plasmodium falciparum infection among school-aged children and to undertake a model-based spatial prediction of P. falciparum infection risk using environmental predictors. A cross-sectional survey was conducted, including parasitological examinations and interviews with more than 5,000 children from 93 schools across Côte d'Ivoire. A finger-prick blood sample was obtained from each child to determine Plasmodium species-specific infection and parasitaemia using Giemsa-stained thick and thin blood films. Household socioeconomic status was assessed through asset ownership and household characteristics. Children were interviewed for preventive measures against malaria. Environmental data were gathered from satellite images and digitized maps. A Bayesian geostatistical stochastic search variable selection procedure was employed to identify factors related to P. falciparum infection risk. Bayesian geostatistical logistic regression models were used to map the spatial distribution of P. falciparum infection and to predict the infection prevalence at non-sampled locations via Bayesian kriging. Complete data sets were available from 5,322 children aged 5-16 years across Côte d'Ivoire. P. falciparum was the predominant species (94.5 %). The Bayesian geostatistical variable selection procedure identified land cover and socioeconomic status as important predictors for infection risk with P. falciparum. Model-based prediction identified high P. falciparum infection risk in the north, central-east, south-east, west and south-west of Côte d'Ivoire. Low-risk areas were found in the south-eastern area close to Abidjan and the south-central and west-central part of the country. The P. falciparum infection risk and related uncertainty estimates for school-aged children in Côte d'Ivoire represent the most up-to-date malaria risk maps. These tools can be used for spatial targeting of malaria control interventions.

Bayesian Lagrangian Data Assimilation and Drifter Deployment Strategies

NASA Astrophysics Data System (ADS)

Dutt, A.; Lermusiaux, P. F. J.

2017-12-01

Ocean currents transport a variety of natural (e.g. water masses, phytoplankton, zooplankton, sediments, etc.) and man-made materials and other objects (e.g. pollutants, floating debris, search and rescue, etc.). Lagrangian Coherent Structures (LCSs) or the most influential/persistent material lines in a flow, provide a robust approach to characterize such Lagrangian transports and organize classic trajectories. Using the flow-map stochastic advection and a dynamically-orthogonal decomposition, we develop uncertainty prediction schemes for both Eulerian and Lagrangian variables. We then extend our Bayesian Gaussian Mixture Model (GMM)-DO filter to a joint Eulerian-Lagrangian Bayesian data assimilation scheme. The resulting nonlinear filter allows the simultaneous non-Gaussian estimation of Eulerian variables (e.g. velocity, temperature, salinity, etc.) and Lagrangian variables (e.g. drifter/float positions, trajectories, LCSs, etc.). Its results are showcased using a double-gyre flow with a random frequency, a stochastic flow past a cylinder, and realistic ocean examples. We further show how our Bayesian mutual information and adaptive sampling equations provide a rigorous efficient methodology to plan optimal drifter deployment strategies and predict the optimal times, locations, and types of measurements to be collected.

Bayesian Genomic Prediction with Genotype × Environment Interaction Kernel Models

PubMed Central

Cuevas, Jaime; Crossa, José; Montesinos-López, Osval A.; Burgueño, Juan; Pérez-Rodríguez, Paulino; de los Campos, Gustavo

2016-01-01

The phenomenon of genotype × environment (G × E) interaction in plant breeding decreases selection accuracy, thereby negatively affecting genetic gains. Several genomic prediction models incorporating G × E have been recently developed and used in genomic selection of plant breeding programs. Genomic prediction models for assessing multi-environment G × E interaction are extensions of a single-environment model, and have advantages and limitations. In this study, we propose two multi-environment Bayesian genomic models: the first model considers genetic effects (u) that can be assessed by the Kronecker product of variance–covariance matrices of genetic correlations between environments and genomic kernels through markers under two linear kernel methods, linear (genomic best linear unbiased predictors, GBLUP) and Gaussian (Gaussian kernel, GK). The other model has the same genetic component as the first model (u) plus an extra component, f, that captures random effects between environments that were not captured by the random effects u. We used five CIMMYT data sets (one maize and four wheat) that were previously used in different studies. Results show that models with G × E always have superior prediction ability than single-environment models, and the higher prediction ability of multi-environment models with u and f over the multi-environment model with only u occurred 85% of the time with GBLUP and 45% of the time with GK across the five data sets. The latter result indicated that including the random effect f is still beneficial for increasing prediction ability after adjusting by the random effect u. PMID:27793970

Bayesian Genomic Prediction with Genotype × Environment Interaction Kernel Models.

PubMed

Cuevas, Jaime; Crossa, José; Montesinos-López, Osval A; Burgueño, Juan; Pérez-Rodríguez, Paulino; de Los Campos, Gustavo

2017-01-05

The phenomenon of genotype × environment (G × E) interaction in plant breeding decreases selection accuracy, thereby negatively affecting genetic gains. Several genomic prediction models incorporating G × E have been recently developed and used in genomic selection of plant breeding programs. Genomic prediction models for assessing multi-environment G × E interaction are extensions of a single-environment model, and have advantages and limitations. In this study, we propose two multi-environment Bayesian genomic models: the first model considers genetic effects [Formula: see text] that can be assessed by the Kronecker product of variance-covariance matrices of genetic correlations between environments and genomic kernels through markers under two linear kernel methods, linear (genomic best linear unbiased predictors, GBLUP) and Gaussian (Gaussian kernel, GK). The other model has the same genetic component as the first model [Formula: see text] plus an extra component, F: , that captures random effects between environments that were not captured by the random effects [Formula: see text] We used five CIMMYT data sets (one maize and four wheat) that were previously used in different studies. Results show that models with G × E always have superior prediction ability than single-environment models, and the higher prediction ability of multi-environment models with [Formula: see text] over the multi-environment model with only u occurred 85% of the time with GBLUP and 45% of the time with GK across the five data sets. The latter result indicated that including the random effect f is still beneficial for increasing prediction ability after adjusting by the random effect [Formula: see text]. Copyright © 2017 Cuevas et al.

Improvement of Storm Forecasts Using Gridded Bayesian Linear Regression for Northeast United States

NASA Astrophysics Data System (ADS)

Yang, J.; Astitha, M.; Schwartz, C. S.

2017-12-01

Bayesian linear regression (BLR) is a post-processing technique in which regression coefficients are derived and used to correct raw forecasts based on pairs of observation-model values. This study presents the development and application of a gridded Bayesian linear regression (GBLR) as a new post-processing technique to improve numerical weather prediction (NWP) of rain and wind storm forecasts over northeast United States. Ten controlled variables produced from ten ensemble members of the National Center for Atmospheric Research (NCAR) real-time prediction system are used for a GBLR model. In the GBLR framework, leave-one-storm-out cross-validation is utilized to study the performances of the post-processing technique in a database composed of 92 storms. To estimate the regression coefficients of the GBLR, optimization procedures that minimize the systematic and random error of predicted atmospheric variables (wind speed, precipitation, etc.) are implemented for the modeled-observed pairs of training storms. The regression coefficients calculated for meteorological stations of the National Weather Service are interpolated back to the model domain. An analysis of forecast improvements based on error reductions during the storms will demonstrate the value of GBLR approach. This presentation will also illustrate how the variances are optimized for the training partition in GBLR and discuss the verification strategy for grid points where no observations are available. The new post-processing technique is successful in improving wind speed and precipitation storm forecasts using past event-based data and has the potential to be implemented in real-time.

An efficient Bayesian data-worth analysis using a multilevel Monte Carlo method

NASA Astrophysics Data System (ADS)

Lu, Dan; Ricciuto, Daniel; Evans, Katherine

2018-03-01

Improving the understanding of subsurface systems and thus reducing prediction uncertainty requires collection of data. As the collection of subsurface data is costly, it is important that the data collection scheme is cost-effective. Design of a cost-effective data collection scheme, i.e., data-worth analysis, requires quantifying model parameter, prediction, and both current and potential data uncertainties. Assessment of these uncertainties in large-scale stochastic subsurface hydrological model simulations using standard Monte Carlo (MC) sampling or surrogate modeling is extremely computationally intensive, sometimes even infeasible. In this work, we propose an efficient Bayesian data-worth analysis using a multilevel Monte Carlo (MLMC) method. Compared to the standard MC that requires a significantly large number of high-fidelity model executions to achieve a prescribed accuracy in estimating expectations, the MLMC can substantially reduce computational costs using multifidelity approximations. Since the Bayesian data-worth analysis involves a great deal of expectation estimation, the cost saving of the MLMC in the assessment can be outstanding. While the proposed MLMC-based data-worth analysis is broadly applicable, we use it for a highly heterogeneous two-phase subsurface flow simulation to select an optimal candidate data set that gives the largest uncertainty reduction in predicting mass flow rates at four production wells. The choices made by the MLMC estimation are validated by the actual measurements of the potential data, and consistent with the standard MC estimation. But compared to the standard MC, the MLMC greatly reduces the computational costs.

Linking big models to big data: efficient ecosystem model calibration through Bayesian model emulation

NASA Astrophysics Data System (ADS)

Fer, I.; Kelly, R.; Andrews, T.; Dietze, M.; Richardson, A. D.

2016-12-01

Our ability to forecast ecosystems is limited by how well we parameterize ecosystem models. Direct measurements for all model parameters are not always possible and inverse estimation of these parameters through Bayesian methods is computationally costly. A solution to computational challenges of Bayesian calibration is to approximate the posterior probability surface using a Gaussian Process that emulates the complex process-based model. Here we report the integration of this method within an ecoinformatics toolbox, Predictive Ecosystem Analyzer (PEcAn), and its application with two ecosystem models: SIPNET and ED2.1. SIPNET is a simple model, allowing application of MCMC methods both to the model itself and to its emulator. We used both approaches to assimilate flux (CO2 and latent heat), soil respiration, and soil carbon data from Bartlett Experimental Forest. This comparison showed that emulator is reliable in terms of convergence to the posterior distribution. A 10000-iteration MCMC analysis with SIPNET itself required more than two orders of magnitude greater computation time than an MCMC run of same length with its emulator. This difference would be greater for a more computationally demanding model. Validation of the emulator-calibrated SIPNET against both the assimilated data and out-of-sample data showed improved fit and reduced uncertainty around model predictions. We next applied the validated emulator method to the ED2, whose complexity precludes standard Bayesian data assimilation. We used the ED2 emulator to assimilate demographic data from a network of inventory plots. For validation of the calibrated ED2, we compared the model to results from Empirical Succession Mapping (ESM), a novel synthesis of successional patterns in Forest Inventory and Analysis data. Our results revealed that while the pre-assimilation ED2 formulation cannot capture the emergent demographic patterns from ESM analysis, constrained model parameters controlling demographic processes increased their agreement considerably.

Simultaneous Force Regression and Movement Classification of Fingers via Surface EMG within a Unified Bayesian Framework.

PubMed

Baldacchino, Tara; Jacobs, William R; Anderson, Sean R; Worden, Keith; Rowson, Jennifer

2018-01-01

This contribution presents a novel methodology for myolectric-based control using surface electromyographic (sEMG) signals recorded during finger movements. A multivariate Bayesian mixture of experts (MoE) model is introduced which provides a powerful method for modeling force regression at the fingertips, while also performing finger movement classification as a by-product of the modeling algorithm. Bayesian inference of the model allows uncertainties to be naturally incorporated into the model structure. This method is tested using data from the publicly released NinaPro database which consists of sEMG recordings for 6 degree-of-freedom force activations for 40 intact subjects. The results demonstrate that the MoE model achieves similar performance compared to the benchmark set by the authors of NinaPro for finger force regression. Additionally, inherent to the Bayesian framework is the inclusion of uncertainty in the model parameters, naturally providing confidence bounds on the force regression predictions. Furthermore, the integrated clustering step allows a detailed investigation into classification of the finger movements, without incurring any extra computational effort. Subsequently, a systematic approach to assessing the importance of the number of electrodes needed for accurate control is performed via sensitivity analysis techniques. A slight degradation in regression performance is observed for a reduced number of electrodes, while classification performance is unaffected.

A Bayes linear Bayes method for estimation of correlated event rates.

PubMed

Quigley, John; Wilson, Kevin J; Walls, Lesley; Bedford, Tim

2013-12-01

Typically, full Bayesian estimation of correlated event rates can be computationally challenging since estimators are intractable. When estimation of event rates represents one activity within a larger modeling process, there is an incentive to develop more efficient inference than provided by a full Bayesian model. We develop a new subjective inference method for correlated event rates based on a Bayes linear Bayes model under the assumption that events are generated from a homogeneous Poisson process. To reduce the elicitation burden we introduce homogenization factors to the model and, as an alternative to a subjective prior, an empirical method using the method of moments is developed. Inference under the new method is compared against estimates obtained under a full Bayesian model, which takes a multivariate gamma prior, where the predictive and posterior distributions are derived in terms of well-known functions. The mathematical properties of both models are presented. A simulation study shows that the Bayes linear Bayes inference method and the full Bayesian model provide equally reliable estimates. An illustrative example, motivated by a problem of estimating correlated event rates across different users in a simple supply chain, shows how ignoring the correlation leads to biased estimation of event rates. © 2013 Society for Risk Analysis.

Quantifying uncertainty in Bayesian calibrated animal-to-human PBPK models with informative prior distributions

EPA Science Inventory

Understanding and quantifying the uncertainty of model parameters and predictions has gained more interest in recent years with the increased use of computational models in chemical risk assessment. Fully characterizing the uncertainty in risk metrics derived from linked quantita...

A probabilistic process model for pelagic marine ecosystems informed by Bayesian inverse analysis

EPA Science Inventory

Marine ecosystems are complex systems with multiple pathways that produce feedback cycles, which may lead to unanticipated effects. Models abstract this complexity and allow us to predict, understand, and hypothesize. In ecological models, however, the paucity of empirical data...

Data-driven Modeling of Metal-oxide Sensors with Dynamic Bayesian Networks

NASA Astrophysics Data System (ADS)

Gosangi, Rakesh; Gutierrez-Osuna, Ricardo

2011-09-01

We present a data-driven probabilistic framework to model the transient response of MOX sensors modulated with a sequence of voltage steps. Analytical models of MOX sensors are usually built based on the physico-chemical properties of the sensing materials. Although building these models provides an insight into the sensor behavior, they also require a thorough understanding of the underlying operating principles. Here we propose a data-driven approach to characterize the dynamical relationship between sensor inputs and outputs. Namely, we use dynamic Bayesian networks (DBNs), probabilistic models that represent temporal relations between a set of random variables. We identify a set of control variables that influence the sensor responses, create a graphical representation that captures the causal relations between these variables, and finally train the model with experimental data. We validated the approach on experimental data in terms of predictive accuracy and classification performance. Our results show that DBNs can accurately predict the dynamic response of MOX sensors, as well as capture the discriminatory information present in the sensor transients.

Tracing the influence of land-use change on water quality and coral reefs using a Bayesian model.

PubMed

Brown, Christopher J; Jupiter, Stacy D; Albert, Simon; Klein, Carissa J; Mangubhai, Sangeeta; Maina, Joseph M; Mumby, Peter; Olley, Jon; Stewart-Koster, Ben; Tulloch, Vivitskaia; Wenger, Amelia

2017-07-06

Coastal ecosystems can be degraded by poor water quality. Tracing the causes of poor water quality back to land-use change is necessary to target catchment management for coastal zone management. However, existing models for tracing the sources of pollution require extensive data-sets which are not available for many of the world's coral reef regions that may have severe water quality issues. Here we develop a hierarchical Bayesian model that uses freely available satellite data to infer the connection between land-uses in catchments and water clarity in coastal oceans. We apply the model to estimate the influence of land-use change on water clarity in Fiji. We tested the model's predictions against underwater surveys, finding that predictions of poor water quality are consistent with observations of high siltation and low coverage of sediment-sensitive coral genera. The model thus provides a means to link land-use change to declines in coastal water quality.

Bayesian Approach for Flexible Modeling of Semicompeting Risks Data

PubMed Central

Han, Baoguang; Yu, Menggang; Dignam, James J.; Rathouz, Paul J.

2016-01-01

Summary Semicompeting risks data arise when two types of events, non-terminal and terminal, are observed. When the terminal event occurs first, it censors the non-terminal event, but not vice versa. To account for possible dependent censoring of the non-terminal event by the terminal event and to improve prediction of the terminal event using the non-terminal event information, it is crucial to model their association properly. Motivated by a breast cancer clinical trial data analysis, we extend the well-known illness-death models to allow flexible random effects to capture heterogeneous association structures in the data. Our extension also represents a generalization of the popular shared frailty models that usually assume that the non-terminal event does not affect the hazards of the terminal event beyond a frailty term. We propose a unified Bayesian modeling approach that can utilize existing software packages for both model fitting and individual specific event prediction. The approach is demonstrated via both simulation studies and a breast cancer data set analysis. PMID:25274445

Bayesian randomized clinical trials: From fixed to adaptive design.

PubMed

Yin, Guosheng; Lam, Chi Kin; Shi, Haolun

2017-08-01

Randomized controlled studies are the gold standard for phase III clinical trials. Using α-spending functions to control the overall type I error rate, group sequential methods are well established and have been dominating phase III studies. Bayesian randomized design, on the other hand, can be viewed as a complement instead of competitive approach to the frequentist methods. For the fixed Bayesian design, the hypothesis testing can be cast in the posterior probability or Bayes factor framework, which has a direct link to the frequentist type I error rate. Bayesian group sequential design relies upon Bayesian decision-theoretic approaches based on backward induction, which is often computationally intensive. Compared with the frequentist approaches, Bayesian methods have several advantages. The posterior predictive probability serves as a useful and convenient tool for trial monitoring, and can be updated at any time as the data accrue during the trial. The Bayesian decision-theoretic framework possesses a direct link to the decision making in the practical setting, and can be modeled more realistically to reflect the actual cost-benefit analysis during the drug development process. Other merits include the possibility of hierarchical modeling and the use of informative priors, which would lead to a more comprehensive utilization of information from both historical and longitudinal data. From fixed to adaptive design, we focus on Bayesian randomized controlled clinical trials and make extensive comparisons with frequentist counterparts through numerical studies. Copyright © 2017 Elsevier Inc. All rights reserved.

Combining PubMed knowledge and EHR data to develop a weighted bayesian network for pancreatic cancer prediction.

PubMed

Zhao, Di; Weng, Chunhua

2011-10-01

In this paper, we propose a novel method that combines PubMed knowledge and Electronic Health Records to develop a weighted Bayesian Network Inference (BNI) model for pancreatic cancer prediction. We selected 20 common risk factors associated with pancreatic cancer and used PubMed knowledge to weigh the risk factors. A keyword-based algorithm was developed to extract and classify PubMed abstracts into three categories that represented positive, negative, or neutral associations between each risk factor and pancreatic cancer. Then we designed a weighted BNI model by adding the normalized weights into a conventional BNI model. We used this model to extract the EHR values for patients with or without pancreatic cancer, which then enabled us to calculate the prior probabilities for the 20 risk factors in the BNI. The software iDiagnosis was designed to use this weighted BNI model for predicting pancreatic cancer. In an evaluation using a case-control dataset, the weighted BNI model significantly outperformed the conventional BNI and two other classifiers (k-Nearest Neighbor and Support Vector Machine). We conclude that the weighted BNI using PubMed knowledge and EHR data shows remarkable accuracy improvement over existing representative methods for pancreatic cancer prediction. Copyright © 2011 Elsevier Inc. All rights reserved.

Combining PubMed Knowledge and EHR Data to Develop a Weighted Bayesian Network for Pancreatic Cancer Prediction

PubMed Central

Zhao, Di; Weng, Chunhua

2011-01-01

In this paper, we propose a novel method that combines PubMed knowledge and Electronic Health Records to develop a weighted Bayesian Network Inference (BNI) model for pancreatic cancer prediction. We selected 20 common risk factors associated with pancreatic cancer and used PubMed knowledge to weigh the risk factors. A keyword-based algorithm was developed to extract and classify PubMed abstracts into three categories that represented positive, negative, or neutral associations between each risk factor and pancreatic cancer. Then we designed a weighted BNI model by adding the normalized weights into a conventional BNI model. We used this model to extract the EHR values for patients with or without pancreatic cancer, which then enabled us to calculate the prior probabilities for the 20 risk factors in the BNI. The software iDiagnosis was designed to use this weighted BNI model for predicting pancreatic cancer. In an evaluation using a case-control dataset, the weighted BNI model significantly outperformed the conventional BNI and two other classifiers (k-Nearest Neighbor and Support Vector Machine). We conclude that the weighted BNI using PubMed knowledge and EHR data shows remarkable accuracy improvement over existing representative methods for pancreatic cancer prediction. PMID:21642013

Response of hydrology to climate change in the southern Appalachian mountains using Bayesian inference

Treesearch

Wei Wu; James S. Clark; James M. Vose

2012-01-01

Predicting long-term consequences of climate change on hydrologic processes has been limited due to the needs to accommodate the uncertainties in hydrological measurements for calibration, and to account for the uncertainties in the models that would ingest those calibrations and uncertainties in climate predictions as basis for hydrological predictions. We implemented...

Uncertainty Quantification of Hypothesis Testing for the Integrated Knowledge Engine

DOE Office of Scientific and Technical Information (OSTI.GOV)

Cuellar, Leticia

2012-05-31

The Integrated Knowledge Engine (IKE) is a tool of Bayesian analysis, based on Bayesian Belief Networks or Bayesian networks for short. A Bayesian network is a graphical model (directed acyclic graph) that allows representing the probabilistic structure of many variables assuming a localized type of dependency called the Markov property. The Markov property in this instance makes any node or random variable to be independent of any non-descendant node given information about its parent. A direct consequence of this property is that it is relatively easy to incorporate new evidence and derive the appropriate consequences, which in general is notmore » an easy or feasible task. Typically we use Bayesian networks as predictive models for a small subset of the variables, either the leave nodes or the root nodes. In IKE, since most applications deal with diagnostics, we are interested in predicting the likelihood of the root nodes given new observations on any of the children nodes. The root nodes represent the various possible outcomes of the analysis, and an important problem is to determine when we have gathered enough evidence to lean toward one of these particular outcomes. This document presents criteria to decide when the evidence gathered is sufficient to draw a particular conclusion or decide in favor of a particular outcome by quantifying the uncertainty in the conclusions that are drawn from the data. The material in this document is organized as follows: Section 2 presents briefly a forensics Bayesian network, and we explore evaluating the information provided by new evidence by looking first at the posterior distribution of the nodes of interest, and then at the corresponding posterior odds ratios. Section 3 presents a third alternative: Bayes Factors. In section 4 we finalize by showing the relation between the posterior odds ratios and Bayes factors and showing examples these cases, and in section 5 we conclude by providing clear guidelines of how to use these for the type of Bayesian networks used in IKE.« less

Parameterization of aquatic ecosystem functioning and its natural variation: Hierarchical Bayesian modelling of plankton food web dynamics

NASA Astrophysics Data System (ADS)

Norros, Veera; Laine, Marko; Lignell, Risto; Thingstad, Frede

2017-10-01

Methods for extracting empirically and theoretically sound parameter values are urgently needed in aquatic ecosystem modelling to describe key flows and their variation in the system. Here, we compare three Bayesian formulations for mechanistic model parameterization that differ in their assumptions about the variation in parameter values between various datasets: 1) global analysis - no variation, 2) separate analysis - independent variation and 3) hierarchical analysis - variation arising from a shared distribution defined by hyperparameters. We tested these methods, using computer-generated and empirical data, coupled with simplified and reasonably realistic plankton food web models, respectively. While all methods were adequate, the simulated example demonstrated that a well-designed hierarchical analysis can result in the most accurate and precise parameter estimates and predictions, due to its ability to combine information across datasets. However, our results also highlighted sensitivity to hyperparameter prior distributions as an important caveat of hierarchical analysis. In the more complex empirical example, hierarchical analysis was able to combine precise identification of parameter values with reasonably good predictive performance, although the ranking of the methods was less straightforward. We conclude that hierarchical Bayesian analysis is a promising tool for identifying key ecosystem-functioning parameters and their variation from empirical datasets.

Probabilistic Space Weather Forecasting: a Bayesian Perspective

NASA Astrophysics Data System (ADS)

Camporeale, E.; Chandorkar, M.; Borovsky, J.; Care', A.

2017-12-01

Most of the Space Weather forecasts, both at operational and research level, are not probabilistic in nature. Unfortunately, a prediction that does not provide a confidence level is not very useful in a decision-making scenario. Nowadays, forecast models range from purely data-driven, machine learning algorithms, to physics-based approximation of first-principle equations (and everything that sits in between). Uncertainties pervade all such models, at every level: from the raw data to finite-precision implementation of numerical methods. The most rigorous way of quantifying the propagation of uncertainties is by embracing a Bayesian probabilistic approach. One of the simplest and most robust machine learning technique in the Bayesian framework is Gaussian Process regression and classification. Here, we present the application of Gaussian Processes to the problems of the DST geomagnetic index forecast, the solar wind type classification, and the estimation of diffusion parameters in radiation belt modeling. In each of these very diverse problems, the GP approach rigorously provide forecasts in the form of predictive distributions. In turn, these distributions can be used as input for ensemble simulations in order to quantify the amplification of uncertainties. We show that we have achieved excellent results in all of the standard metrics to evaluate our models, with very modest computational cost.

Bayesian Cue Integration as a Developmental Outcome of Reward Mediated Learning

PubMed Central

Weisswange, Thomas H.; Rothkopf, Constantin A.; Rodemann, Tobias; Triesch, Jochen

2011-01-01

Average human behavior in cue combination tasks is well predicted by Bayesian inference models. As this capability is acquired over developmental timescales, the question arises, how it is learned. Here we investigated whether reward dependent learning, that is well established at the computational, behavioral, and neuronal levels, could contribute to this development. It is shown that a model free reinforcement learning algorithm can indeed learn to do cue integration, i.e. weight uncertain cues according to their respective reliabilities and even do so if reliabilities are changing. We also consider the case of causal inference where multimodal signals can originate from one or multiple separate objects and should not always be integrated. In this case, the learner is shown to develop a behavior that is closest to Bayesian model averaging. We conclude that reward mediated learning could be a driving force for the development of cue integration and causal inference. PMID:21750717

A Bayesian Account of Vocal Adaptation to Pitch-Shifted Auditory Feedback

PubMed Central

Hahnloser, Richard H. R.

2017-01-01

Motor systems are highly adaptive. Both birds and humans compensate for synthetically induced shifts in the pitch (fundamental frequency) of auditory feedback stemming from their vocalizations. Pitch-shift compensation is partial in the sense that large shifts lead to smaller relative compensatory adjustments of vocal pitch than small shifts. Also, compensation is larger in subjects with high motor variability. To formulate a mechanistic description of these findings, we adapt a Bayesian model of error relevance. We assume that vocal-auditory feedback loops in the brain cope optimally with known sensory and motor variability. Based on measurements of motor variability, optimal compensatory responses in our model provide accurate fits to published experimental data. Optimal compensation correctly predicts sensory acuity, which has been estimated in psychophysical experiments as just-noticeable pitch differences. Our model extends the utility of Bayesian approaches to adaptive vocal behaviors. PMID:28135267

Efficient Probabilistic Diagnostics for Electrical Power Systems

NASA Technical Reports Server (NTRS)

Mengshoel, Ole J.; Chavira, Mark; Cascio, Keith; Poll, Scott; Darwiche, Adnan; Uckun, Serdar

2008-01-01

We consider in this work the probabilistic approach to model-based diagnosis when applied to electrical power systems (EPSs). Our probabilistic approach is formally well-founded, as it based on Bayesian networks and arithmetic circuits. We investigate the diagnostic task known as fault isolation, and pay special attention to meeting two of the main challenges . model development and real-time reasoning . often associated with real-world application of model-based diagnosis technologies. To address the challenge of model development, we develop a systematic approach to representing electrical power systems as Bayesian networks, supported by an easy-to-use speci.cation language. To address the real-time reasoning challenge, we compile Bayesian networks into arithmetic circuits. Arithmetic circuit evaluation supports real-time diagnosis by being predictable and fast. In essence, we introduce a high-level EPS speci.cation language from which Bayesian networks that can diagnose multiple simultaneous failures are auto-generated, and we illustrate the feasibility of using arithmetic circuits, compiled from Bayesian networks, for real-time diagnosis on real-world EPSs of interest to NASA. The experimental system is a real-world EPS, namely the Advanced Diagnostic and Prognostic Testbed (ADAPT) located at the NASA Ames Research Center. In experiments with the ADAPT Bayesian network, which currently contains 503 discrete nodes and 579 edges, we .nd high diagnostic accuracy in scenarios where one to three faults, both in components and sensors, were inserted. The time taken to compute the most probable explanation using arithmetic circuits has a small mean of 0.2625 milliseconds and standard deviation of 0.2028 milliseconds. In experiments with data from ADAPT we also show that arithmetic circuit evaluation substantially outperforms joint tree propagation and variable elimination, two alternative algorithms for diagnosis using Bayesian network inference.

Bayesian predictive power: choice of prior and some recommendations for its use as probability of success in drug development.

PubMed

Rufibach, Kaspar; Burger, Hans Ulrich; Abt, Markus

2016-09-01

Bayesian predictive power, the expectation of the power function with respect to a prior distribution for the true underlying effect size, is routinely used in drug development to quantify the probability of success of a clinical trial. Choosing the prior is crucial for the properties and interpretability of Bayesian predictive power. We review recommendations on the choice of prior for Bayesian predictive power and explore its features as a function of the prior. The density of power values induced by a given prior is derived analytically and its shape characterized. We find that for a typical clinical trial scenario, this density has a u-shape very similar, but not equal, to a β-distribution. Alternative priors are discussed, and practical recommendations to assess the sensitivity of Bayesian predictive power to its input parameters are provided. Copyright © 2016 John Wiley & Sons, Ltd. Copyright © 2016 John Wiley & Sons, Ltd.

Bayesian Analysis of the Association between Family-Level Factors and Siblings' Dental Caries.

PubMed

Wen, A; Weyant, R J; McNeil, D W; Crout, R J; Neiswanger, K; Marazita, M L; Foxman, B

2017-07-01

We conducted a Bayesian analysis of the association between family-level socioeconomic status and smoking and the prevalence of dental caries among siblings (children from infant to 14 y) among children living in rural and urban Northern Appalachia using data from the Center for Oral Health Research in Appalachia (COHRA). The observed proportion of siblings sharing caries was significantly different from predicted assuming siblings' caries status was independent. Using a Bayesian hierarchical model, we found the inclusion of a household factor significantly improved the goodness of fit. Other findings showed an inverse association between parental education and siblings' caries and a positive association between households with smokers and siblings' caries. Our study strengthens existing evidence suggesting that increased parental education and decreased parental cigarette smoking are associated with reduced childhood caries in the household. Our results also demonstrate the value of a Bayesian approach, which allows us to include household as a random effect, thereby providing more accurate estimates than obtained using generalized linear mixed models.

Dimensionality of the 9-item Utrecht Work Engagement Scale revisited: A Bayesian structural equation modeling approach.

PubMed

Fong, Ted C T; Ho, Rainbow T H

2015-01-01

The aim of this study was to reexamine the dimensionality of the widely used 9-item Utrecht Work Engagement Scale using the maximum likelihood (ML) approach and Bayesian structural equation modeling (BSEM) approach. Three measurement models (1-factor, 3-factor, and bi-factor models) were evaluated in two split samples of 1,112 health-care workers using confirmatory factor analysis and BSEM, which specified small-variance informative priors for cross-loadings and residual covariances. Model fit and comparisons were evaluated by posterior predictive p-value (PPP), deviance information criterion, and Bayesian information criterion (BIC). None of the three ML-based models showed an adequate fit to the data. The use of informative priors for cross-loadings did not improve the PPP for the models. The 1-factor BSEM model with approximately zero residual covariances displayed a good fit (PPP>0.10) to both samples and a substantially lower BIC than its 3-factor and bi-factor counterparts. The BSEM results demonstrate empirical support for the 1-factor model as a parsimonious and reasonable representation of work engagement.

Clinical Outcome Prediction in Aneurysmal Subarachnoid Hemorrhage Using Bayesian Neural Networks with Fuzzy Logic Inferences

PubMed Central

Lo, Benjamin W. Y.; Macdonald, R. Loch; Baker, Andrew; Levine, Mitchell A. H.

2013-01-01

Objective. The novel clinical prediction approach of Bayesian neural networks with fuzzy logic inferences is created and applied to derive prognostic decision rules in cerebral aneurysmal subarachnoid hemorrhage (aSAH). Methods. The approach of Bayesian neural networks with fuzzy logic inferences was applied to data from five trials of Tirilazad for aneurysmal subarachnoid hemorrhage (3551 patients). Results. Bayesian meta-analyses of observational studies on aSAH prognostic factors gave generalizable posterior distributions of population mean log odd ratios (ORs). Similar trends were noted in Bayesian and linear regression ORs. Significant outcome predictors include normal motor response, cerebral infarction, history of myocardial infarction, cerebral edema, history of diabetes mellitus, fever on day 8, prior subarachnoid hemorrhage, admission angiographic vasospasm, neurological grade, intraventricular hemorrhage, ruptured aneurysm size, history of hypertension, vasospasm day, age and mean arterial pressure. Heteroscedasticity was present in the nontransformed dataset. Artificial neural networks found nonlinear relationships with 11 hidden variables in 1 layer, using the multilayer perceptron model. Fuzzy logic decision rules (centroid defuzzification technique) denoted cut-off points for poor prognosis at greater than 2.5 clusters. Discussion. This aSAH prognostic system makes use of existing knowledge, recognizes unknown areas, incorporates one's clinical reasoning, and compensates for uncertainty in prognostication. PMID:23690884

Predictability of extreme weather events for NE U.S.: improvement of the numerical prediction using a Bayesian regression approach

NASA Astrophysics Data System (ADS)

Yang, J.; Astitha, M.; Anagnostou, E. N.; Hartman, B.; Kallos, G. B.

2015-12-01

Weather prediction accuracy has become very important for the Northeast U.S. given the devastating effects of extreme weather events in the recent years. Weather forecasting systems are used towards building strategies to prevent catastrophic losses for human lives and the environment. Concurrently, weather forecast tools and techniques have evolved with improved forecast skill as numerical prediction techniques are strengthened by increased super-computing resources. In this study, we examine the combination of two state-of-the-science atmospheric models (WRF and RAMS/ICLAMS) by utilizing a Bayesian regression approach to improve the prediction of extreme weather events for NE U.S. The basic concept behind the Bayesian regression approach is to take advantage of the strengths of two atmospheric modeling systems and, similar to the multi-model ensemble approach, limit their weaknesses which are related to systematic and random errors in the numerical prediction of physical processes. The first part of this study is focused on retrospective simulations of seventeen storms that affected the region in the period 2004-2013. Optimal variances are estimated by minimizing the root mean square error and are applied to out-of-sample weather events. The applicability and usefulness of this approach are demonstrated by conducting an error analysis based on in-situ observations from meteorological stations of the National Weather Service (NWS) for wind speed and wind direction, and NCEP Stage IV radar data, mosaicked from the regional multi-sensor for precipitation. The preliminary results indicate a significant improvement in the statistical metrics of the modeled-observed pairs for meteorological variables using various combinations of the sixteen events as predictors of the seventeenth. This presentation will illustrate the implemented methodology and the obtained results for wind speed, wind direction and precipitation, as well as set the research steps that will be followed in the future.

Learning Parsimonious Classification Rules from Gene Expression Data Using Bayesian Networks with Local Structure.

PubMed

Lustgarten, Jonathan Lyle; Balasubramanian, Jeya Balaji; Visweswaran, Shyam; Gopalakrishnan, Vanathi

2017-03-01

The comprehensibility of good predictive models learned from high-dimensional gene expression data is attractive because it can lead to biomarker discovery. Several good classifiers provide comparable predictive performance but differ in their abilities to summarize the observed data. We extend a Bayesian Rule Learning (BRL-GSS) algorithm, previously shown to be a significantly better predictor than other classical approaches in this domain. It searches a space of Bayesian networks using a decision tree representation of its parameters with global constraints, and infers a set of IF-THEN rules. The number of parameters and therefore the number of rules are combinatorial to the number of predictor variables in the model. We relax these global constraints to a more generalizable local structure (BRL-LSS). BRL-LSS entails more parsimonious set of rules because it does not have to generate all combinatorial rules. The search space of local structures is much richer than the space of global structures. We design the BRL-LSS with the same worst-case time-complexity as BRL-GSS while exploring a richer and more complex model space. We measure predictive performance using Area Under the ROC curve (AUC) and Accuracy. We measure model parsimony performance by noting the average number of rules and variables needed to describe the observed data. We evaluate the predictive and parsimony performance of BRL-GSS, BRL-LSS and the state-of-the-art C4.5 decision tree algorithm, across 10-fold cross-validation using ten microarray gene-expression diagnostic datasets. In these experiments, we observe that BRL-LSS is similar to BRL-GSS in terms of predictive performance, while generating a much more parsimonious set of rules to explain the same observed data. BRL-LSS also needs fewer variables than C4.5 to explain the data with similar predictive performance. We also conduct a feasibility study to demonstrate the general applicability of our BRL methods on the newer RNA sequencing gene-expression data.

Underestimation of Variance of Predicted Health Utilities Derived from Multiattribute Utility Instruments.

PubMed

Chan, Kelvin K W; Xie, Feng; Willan, Andrew R; Pullenayegum, Eleanor M

2017-04-01

Parameter uncertainty in value sets of multiattribute utility-based instruments (MAUIs) has received little attention previously. This false precision leads to underestimation of the uncertainty of the results of cost-effectiveness analyses. The aim of this study is to examine the use of multiple imputation as a method to account for this uncertainty of MAUI scoring algorithms. We fitted a Bayesian model with random effects for respondents and health states to the data from the original US EQ-5D-3L valuation study, thereby estimating the uncertainty in the EQ-5D-3L scoring algorithm. We applied these results to EQ-5D-3L data from the Commonwealth Fund (CWF) Survey for Sick Adults ( n = 3958), comparing the standard error of the estimated mean utility in the CWF population using the predictive distribution from the Bayesian mixed-effect model (i.e., incorporating parameter uncertainty in the value set) with the standard error of the estimated mean utilities based on multiple imputation and the standard error using the conventional approach of using MAUI (i.e., ignoring uncertainty in the value set). The mean utility in the CWF population based on the predictive distribution of the Bayesian model was 0.827 with a standard error (SE) of 0.011. When utilities were derived using the conventional approach, the estimated mean utility was 0.827 with an SE of 0.003, which is only 25% of the SE based on the full predictive distribution of the mixed-effect model. Using multiple imputation with 20 imputed sets, the mean utility was 0.828 with an SE of 0.011, which is similar to the SE based on the full predictive distribution. Ignoring uncertainty of the predicted health utilities derived from MAUIs could lead to substantial underestimation of the variance of mean utilities. Multiple imputation corrects for this underestimation so that the results of cost-effectiveness analyses using MAUIs can report the correct degree of uncertainty.

Two-Stage Bayesian Model Averaging in Endogenous Variable Models*

PubMed Central

Lenkoski, Alex; Eicher, Theo S.; Raftery, Adrian E.

2013-01-01

Economic modeling in the presence of endogeneity is subject to model uncertainty at both the instrument and covariate level. We propose a Two-Stage Bayesian Model Averaging (2SBMA) methodology that extends the Two-Stage Least Squares (2SLS) estimator. By constructing a Two-Stage Unit Information Prior in the endogenous variable model, we are able to efficiently combine established methods for addressing model uncertainty in regression models with the classic technique of 2SLS. To assess the validity of instruments in the 2SBMA context, we develop Bayesian tests of the identification restriction that are based on model averaged posterior predictive p-values. A simulation study showed that 2SBMA has the ability to recover structure in both the instrument and covariate set, and substantially improves the sharpness of resulting coefficient estimates in comparison to 2SLS using the full specification in an automatic fashion. Due to the increased parsimony of the 2SBMA estimate, the Bayesian Sargan test had a power of 50 percent in detecting a violation of the exogeneity assumption, while the method based on 2SLS using the full specification had negligible power. We apply our approach to the problem of development accounting, and find support not only for institutions, but also for geography and integration as development determinants, once both model uncertainty and endogeneity have been jointly addressed. PMID:24223471

Distributed multisensory integration in a recurrent network model through supervised learning

NASA Astrophysics Data System (ADS)

Wang, He; Wong, K. Y. Michael

Sensory integration between different modalities has been extensively studied. It is suggested that the brain integrates signals from different modalities in a Bayesian optimal way. However, how the Bayesian rule is implemented in a neural network remains under debate. In this work we propose a biologically plausible recurrent network model, which can perform Bayesian multisensory integration after trained by supervised learning. Our model is composed of two modules, each for one modality. We assume that each module is a recurrent network, whose activity represents the posterior distribution of each stimulus. The feedforward input on each module is the likelihood of each modality. Two modules are integrated through cross-links, which are feedforward connections from the other modality, and reciprocal connections, which are recurrent connections between different modules. By stochastic gradient descent, we successfully trained the feedforward and recurrent coupling matrices simultaneously, both of which resembles the Mexican-hat. We also find that there are more than one set of coupling matrices that can approximate the Bayesian theorem well. Specifically, reciprocal connections and cross-links will compensate each other if one of them is removed. Even though trained with two inputs, the network's performance with only one input is in good accordance with what is predicted by the Bayesian theorem.

Multi-Model Ensemble Wake Vortex Prediction

NASA Technical Reports Server (NTRS)

Koerner, Stephan; Holzaepfel, Frank; Ahmad, Nash'at N.

2015-01-01

Several multi-model ensemble methods are investigated for predicting wake vortex transport and decay. This study is a joint effort between National Aeronautics and Space Administration and Deutsches Zentrum fuer Luft- und Raumfahrt to develop a multi-model ensemble capability using their wake models. An overview of different multi-model ensemble methods and their feasibility for wake applications is presented. The methods include Reliability Ensemble Averaging, Bayesian Model Averaging, and Monte Carlo Simulations. The methodologies are evaluated using data from wake vortex field experiments.

Statistical modelling of networked human-automation performance using working memory capacity.

PubMed

Ahmed, Nisar; de Visser, Ewart; Shaw, Tyler; Mohamed-Ameen, Amira; Campbell, Mark; Parasuraman, Raja

2014-01-01

This study examines the challenging problem of modelling the interaction between individual attentional limitations and decision-making performance in networked human-automation system tasks. Analysis of real experimental data from a task involving networked supervision of multiple unmanned aerial vehicles by human participants shows that both task load and network message quality affect performance, but that these effects are modulated by individual differences in working memory (WM) capacity. These insights were used to assess three statistical approaches for modelling and making predictions with real experimental networked supervisory performance data: classical linear regression, non-parametric Gaussian processes and probabilistic Bayesian networks. It is shown that each of these approaches can help designers of networked human-automated systems cope with various uncertainties in order to accommodate future users by linking expected operating conditions and performance from real experimental data to observable cognitive traits like WM capacity. Practitioner Summary: Working memory (WM) capacity helps account for inter-individual variability in operator performance in networked unmanned aerial vehicle supervisory tasks. This is useful for reliable performance prediction near experimental conditions via linear models; robust statistical prediction beyond experimental conditions via Gaussian process models and probabilistic inference about unknown task conditions/WM capacities via Bayesian network models.

BFLCRM: A BAYESIAN FUNCTIONAL LINEAR COX REGRESSION MODEL FOR PREDICTING TIME TO CONVERSION TO ALZHEIMER’S DISEASE*

PubMed Central

Lee, Eunjee; Zhu, Hongtu; Kong, Dehan; Wang, Yalin; Giovanello, Kelly Sullivan; Ibrahim, Joseph G

2015-01-01

The aim of this paper is to develop a Bayesian functional linear Cox regression model (BFLCRM) with both functional and scalar covariates. This new development is motivated by establishing the likelihood of conversion to Alzheimer’s disease (AD) in 346 patients with mild cognitive impairment (MCI) enrolled in the Alzheimer’s Disease Neuroimaging Initiative 1 (ADNI-1) and the early markers of conversion. These 346 MCI patients were followed over 48 months, with 161 MCI participants progressing to AD at 48 months. The functional linear Cox regression model was used to establish that functional covariates including hippocampus surface morphology and scalar covariates including brain MRI volumes, cognitive performance (ADAS-Cog), and APOE status can accurately predict time to onset of AD. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. A simulation study is performed to evaluate the finite sample performance of BFLCRM. PMID:26900412

Bayesian Ensemble Trees (BET) for Clustering and Prediction in Heterogeneous Data

PubMed Central

Duan, Leo L.; Clancy, John P.; Szczesniak, Rhonda D.

2016-01-01

We propose a novel “tree-averaging” model that utilizes the ensemble of classification and regression trees (CART). Each constituent tree is estimated with a subset of similar data. We treat this grouping of subsets as Bayesian Ensemble Trees (BET) and model them as a Dirichlet process. We show that BET determines the optimal number of trees by adapting to the data heterogeneity. Compared with the other ensemble methods, BET requires much fewer trees and shows equivalent prediction accuracy using weighted averaging. Moreover, each tree in BET provides variable selection criterion and interpretation for each subset. We developed an efficient estimating procedure with improved estimation strategies in both CART and mixture models. We demonstrate these advantages of BET with simulations and illustrate the approach with a real-world data example involving regression of lung function measurements obtained from patients with cystic fibrosis. Supplemental materials are available online. PMID:27524872

On the limitations of standard statistical modeling in biological systems: a full Bayesian approach for biology.

PubMed

Gomez-Ramirez, Jaime; Sanz, Ricardo

2013-09-01

One of the most important scientific challenges today is the quantitative and predictive understanding of biological function. Classical mathematical and computational approaches have been enormously successful in modeling inert matter, but they may be inadequate to address inherent features of biological systems. We address the conceptual and methodological obstacles that lie in the inverse problem in biological systems modeling. We introduce a full Bayesian approach (FBA), a theoretical framework to study biological function, in which probability distributions are conditional on biophysical information that physically resides in the biological system that is studied by the scientist. Copyright © 2013 Elsevier Ltd. All rights reserved.

Model estimation of claim risk and premium for motor vehicle insurance by using Bayesian method

NASA Astrophysics Data System (ADS)

Sukono; Riaman; Lesmana, E.; Wulandari, R.; Napitupulu, H.; Supian, S.

2018-01-01

Risk models need to be estimated by the insurance company in order to predict the magnitude of the claim and determine the premiums charged to the insured. This is intended to prevent losses in the future. In this paper, we discuss the estimation of risk model claims and motor vehicle insurance premiums using Bayesian methods approach. It is assumed that the frequency of claims follow a Poisson distribution, while a number of claims assumed to follow a Gamma distribution. The estimation of parameters of the distribution of the frequency and amount of claims are made by using Bayesian methods. Furthermore, the estimator distribution of frequency and amount of claims are used to estimate the aggregate risk models as well as the value of the mean and variance. The mean and variance estimator that aggregate risk, was used to predict the premium eligible to be charged to the insured. Based on the analysis results, it is shown that the frequency of claims follow a Poisson distribution with parameter values λ is 5.827. While a number of claims follow the Gamma distribution with parameter values p is 7.922 and θ is 1.414. Therefore, the obtained values of the mean and variance of the aggregate claims respectively are IDR 32,667,489.88 and IDR 38,453,900,000,000.00. In this paper the prediction of the pure premium eligible charged to the insured is obtained, which amounting to IDR 2,722,290.82. The prediction of the claims and premiums aggregate can be used as a reference for the insurance company’s decision-making in management of reserves and premiums of motor vehicle insurance.

Bayesian Framework Approach for Prognostic Studies in Electrolytic Capacitor under Thermal Overstress Conditions

NASA Technical Reports Server (NTRS)

Kulkarni, Chetan S.; Celaya, Jose R.; Goebel, Kai; Biswas, Gautam

2012-01-01

Electrolytic capacitors are used in several applications ranging from power supplies for safety critical avionics equipment to power drivers for electro-mechanical actuator. Past experiences show that capacitors tend to degrade and fail faster when subjected to high electrical or thermal stress conditions during operations. This makes them good candidates for prognostics and health management. Model-based prognostics captures system knowledge in the form of physics-based models of components in order to obtain accurate predictions of end of life based on their current state of heal th and their anticipated future use and operational conditions. The focus of this paper is on deriving first principles degradation models for thermal stress conditions and implementing Bayesian framework for making remaining useful life predictions. Data collected from simultaneous experiments are used to validate the models. Our overall goal is to derive accurate models of capacitor degradation, and use them to remaining useful life in DC-DC converters.

Uncertainty in the Bayesian meta-analysis of normally distributed surrogate endpoints

PubMed Central

Thompson, John R; Spata, Enti; Abrams, Keith R

2015-01-01

We investigate the effect of the choice of parameterisation of meta-analytic models and related uncertainty on the validation of surrogate endpoints. Different meta-analytical approaches take into account different levels of uncertainty which may impact on the accuracy of the predictions of treatment effect on the target outcome from the treatment effect on a surrogate endpoint obtained from these models. A range of Bayesian as well as frequentist meta-analytical methods are implemented using illustrative examples in relapsing–remitting multiple sclerosis, where the treatment effect on disability worsening is the primary outcome of interest in healthcare evaluation, while the effect on relapse rate is considered as a potential surrogate to the effect on disability progression, and in gastric cancer, where the disease-free survival has been shown to be a good surrogate endpoint to the overall survival. Sensitivity analysis was carried out to assess the impact of distributional assumptions on the predictions. Also, sensitivity to modelling assumptions and performance of the models were investigated by simulation. Although different methods can predict mean true outcome almost equally well, inclusion of uncertainty around all relevant parameters of the model may lead to less certain and hence more conservative predictions. When investigating endpoints as candidate surrogate outcomes, a careful choice of the meta-analytical approach has to be made. Models underestimating the uncertainty of available evidence may lead to overoptimistic predictions which can then have an effect on decisions made based on such predictions. PMID:26271918

Uncertainty in the Bayesian meta-analysis of normally distributed surrogate endpoints.

PubMed

Bujkiewicz, Sylwia; Thompson, John R; Spata, Enti; Abrams, Keith R

2017-10-01

We investigate the effect of the choice of parameterisation of meta-analytic models and related uncertainty on the validation of surrogate endpoints. Different meta-analytical approaches take into account different levels of uncertainty which may impact on the accuracy of the predictions of treatment effect on the target outcome from the treatment effect on a surrogate endpoint obtained from these models. A range of Bayesian as well as frequentist meta-analytical methods are implemented using illustrative examples in relapsing-remitting multiple sclerosis, where the treatment effect on disability worsening is the primary outcome of interest in healthcare evaluation, while the effect on relapse rate is considered as a potential surrogate to the effect on disability progression, and in gastric cancer, where the disease-free survival has been shown to be a good surrogate endpoint to the overall survival. Sensitivity analysis was carried out to assess the impact of distributional assumptions on the predictions. Also, sensitivity to modelling assumptions and performance of the models were investigated by simulation. Although different methods can predict mean true outcome almost equally well, inclusion of uncertainty around all relevant parameters of the model may lead to less certain and hence more conservative predictions. When investigating endpoints as candidate surrogate outcomes, a careful choice of the meta-analytical approach has to be made. Models underestimating the uncertainty of available evidence may lead to overoptimistic predictions which can then have an effect on decisions made based on such predictions.

Using a Data-Driven Approach to Understand the Interaction between Catchment Characteristics and Water Quality Responses

NASA Astrophysics Data System (ADS)

Western, A. W.; Lintern, A.; Liu, S.; Ryu, D.; Webb, J. A.; Leahy, P.; Wilson, P.; Waters, D.; Bende-Michl, U.; Watson, M.

2016-12-01

Many streams, lakes and estuaries are experiencing increasing concentrations and loads of nutrient and sediments. Models that can predict the spatial and temporal variability in water quality of aquatic systems are required to help guide the management and restoration of polluted aquatic systems. We propose that a Bayesian hierarchical modelling framework could be used to predict water quality responses over varying spatial and temporal scales. Stream water quality data and spatial data of catchment characteristics collected throughout Victoria and Queensland (in Australia) over two decades will be used to develop this Bayesian hierarchical model. In this paper, we present the preliminary exploratory data analysis required for the development of the Bayesian hierarchical model. Specifically, we present the results of exploratory data analysis of Total Nitrogen (TN) concentrations in rivers in Victoria (in South-East Australia) to illustrate the catchment characteristics that appear to be influencing spatial variability in (1) mean concentrations of TN; and (2) the relationship between discharge and TN throughout the state. These important catchment characteristics were identified using: (1) monthly TN concentrations measured at 28 water quality gauging stations and (2) climate, land use, topographic and geologic characteristics of the catchments of these 28 sites. Spatial variability in TN concentrations had a positive correlation to fertiliser use in the catchment and average temperature. There were negative correlations between TN concentrations and catchment forest cover, annual runoff, runoff perenniality, soil erosivity and catchment slope. The relationship between discharge and TN concentrations showed spatial variability, possibly resulting from climatic and topographic differences between the sites. The results of this study will feed into the hierarchical Bayesian model of river water quality.

Bayesian quantitative precipitation forecasts in terms of quantiles

NASA Astrophysics Data System (ADS)

Bentzien, Sabrina; Friederichs, Petra

2014-05-01

Ensemble prediction systems (EPS) for numerical weather predictions on the mesoscale are particularly developed to obtain probabilistic guidance for high impact weather. An EPS not only issues a deterministic future state of the atmosphere but a sample of possible future states. Ensemble postprocessing then translates such a sample of forecasts into probabilistic measures. This study focus on probabilistic quantitative precipitation forecasts in terms of quantiles. Quantiles are particular suitable to describe precipitation at various locations, since no assumption is required on the distribution of precipitation. The focus is on the prediction during high-impact events and related to the Volkswagen Stiftung funded project WEX-MOP (Mesoscale Weather Extremes - Theory, Spatial Modeling and Prediction). Quantile forecasts are derived from the raw ensemble and via quantile regression. Neighborhood method and time-lagging are effective tools to inexpensively increase the ensemble spread, which results in more reliable forecasts especially for extreme precipitation events. Since an EPS provides a large amount of potentially informative predictors, a variable selection is required in order to obtain a stable statistical model. A Bayesian formulation of quantile regression allows for inference about the selection of predictive covariates by the use of appropriate prior distributions. Moreover, the implementation of an additional process layer for the regression parameters accounts for spatial variations of the parameters. Bayesian quantile regression and its spatially adaptive extension is illustrated for the German-focused mesoscale weather prediction ensemble COSMO-DE-EPS, which runs (pre)operationally since December 2010 at the German Meteorological Service (DWD). Objective out-of-sample verification uses the quantile score (QS), a weighted absolute error between quantile forecasts and observations. The QS is a proper scoring function and can be decomposed into reliability, resolutions and uncertainty parts. A quantile reliability plot gives detailed insights in the predictive performance of the quantile forecasts.

OGLE-2008-BLG-355Lb: A massive planet around a late-type star

DOE Office of Scientific and Technical Information (OSTI.GOV)

Koshimoto, N.; Sumi, T.; Fukagawa, M.

2014-06-20

We report the discovery of a massive planet, OGLE-2008-BLG-355Lb. The light curve analysis indicates a planet:host mass ratio of q = 0.0118 ± 0.0006 at a separation of 0.877 ± 0.010 Einstein radii. We do not measure a significant microlensing parallax signal and do not have high angular resolution images that could detect the planetary host star. Therefore, we do not have a direct measurement of the host star mass. A Bayesian analysis, assuming that all host stars have equal probability to host a planet with the measured mass ratio, implies a host star mass of M{sub h}=0.37{sub −0.17}{sup +0.30}more » M{sub ⊙} and a companion of mass M{sub P}=4.6{sub −2.2}{sup +3.7}M{sub J}, at a projected separation of r{sub ⊥}=1.70{sub −0.30}{sup +0.29} AU. The implied distance to the planetary system is D {sub L} = 6.8 ± 1.1 kpc. A planetary system with the properties preferred by the Bayesian analysis may be a challenge to the core accretion model of planet formation, as the core accretion model predicts that massive planets are far more likely to form around more massive host stars. This core accretion model prediction is not consistent with our Bayesian prior of an equal probability of host stars of all masses to host a planet with the measured mass ratio. Thus, if the core accretion model prediction is right, we should expect that follow-up high angular resolution observations will detect a host star with a mass in the upper part of the range allowed by the Bayesian analysis. That is, the host would probably be a K or G dwarf.« less

Bayesian Evaluation of Dynamical Soil Carbon Models Using Soil Carbon Flux Data

NASA Astrophysics Data System (ADS)

Xie, H. W.; Romero-Olivares, A.; Guindani, M.; Allison, S. D.

2017-12-01

2016 was Earth's hottest year in the modern temperature record and the third consecutive record-breaking year. As the planet continues to warm, temperature-induced changes in respiration rates of soil microbes could reduce the amount of carbon sequestered in the soil organic carbon (SOC) pool, one of the largest terrestrial stores of carbon. This would accelerate temperature increases. In order to predict the future size of the SOC pool, mathematical soil carbon models (SCMs) describing interactions between the biosphere and atmosphere are needed. SCMs must be validated before they can be chosen for predictive use. In this study, we check two SCMs called CON and AWB for consistency with observed data using Bayesian goodness of fit testing that can be used in the future to compare other models. We compare the fit of the models to longitudinal soil respiration data from a meta-analysis of soil heating experiments using a family of Bayesian goodness of fit metrics called information criteria (IC), including the Widely Applicable Information Criterion (WAIC), the Leave-One-Out Information Criterion (LOOIC), and the Log Pseudo Marginal Likelihood (LPML). These IC's take the entire posterior distribution into account, rather than just one outputted model fit line. A lower WAIC and LOOIC and larger LPML indicate a better fit. We compare AWB and CON with fixed steady state model pool sizes. At equivalent SOC, dissolved organic carbon, and microbial pool sizes, CON always outperforms AWB quantitatively by all three IC's used. AWB monotonically improves in fit as we reduce the SOC steady state pool size while fixing all other pool sizes, and the same is almost true for CON. The AWB model with the lowest SOC is the best performing AWB model, while the CON model with the second lowest SOC is the best performing model. We observe that AWB displays more changes in slope sign and qualitatively displays more adaptive dynamics, which prevents AWB from being fully ruled out for predictive use, but based on IC's, CON is clearly the superior model for fitting the data. Hence, we demonstrate that Bayesian goodness of fit testing with information criteria helps us rigorously determine the consistency of models with data. Models that demonstrate their consistency to multiple data sets with our approach can then be selected for further refinement.

A hierarchical Bayesian GEV model for improving local and regional flood quantile estimates

NASA Astrophysics Data System (ADS)

Lima, Carlos H. R.; Lall, Upmanu; Troy, Tara; Devineni, Naresh

2016-10-01

We estimate local and regional Generalized Extreme Value (GEV) distribution parameters for flood frequency analysis in a multilevel, hierarchical Bayesian framework, to explicitly model and reduce uncertainties. As prior information for the model, we assume that the GEV location and scale parameters for each site come from independent log-normal distributions, whose mean parameter scales with the drainage area. From empirical and theoretical arguments, the shape parameter for each site is shrunk towards a common mean. Non-informative prior distributions are assumed for the hyperparameters and the MCMC method is used to sample from the joint posterior distribution. The model is tested using annual maximum series from 20 streamflow gauges located in an 83,000 km2 flood prone basin in Southeast Brazil. The results show a significant reduction of uncertainty estimates of flood quantile estimates over the traditional GEV model, particularly for sites with shorter records. For return periods within the range of the data (around 50 years), the Bayesian credible intervals for the flood quantiles tend to be narrower than the classical confidence limits based on the delta method. As the return period increases beyond the range of the data, the confidence limits from the delta method become unreliable and the Bayesian credible intervals provide a way to estimate satisfactory confidence bands for the flood quantiles considering parameter uncertainties and regional information. In order to evaluate the applicability of the proposed hierarchical Bayesian model for regional flood frequency analysis, we estimate flood quantiles for three randomly chosen out-of-sample sites and compare with classical estimates using the index flood method. The posterior distributions of the scaling law coefficients are used to define the predictive distributions of the GEV location and scale parameters for the out-of-sample sites given only their drainage areas and the posterior distribution of the average shape parameter is taken as the regional predictive distribution for this parameter. While the index flood method does not provide a straightforward way to consider the uncertainties in the index flood and in the regional parameters, the results obtained here show that the proposed Bayesian method is able to produce adequate credible intervals for flood quantiles that are in accordance with empirical estimates.

Bayesian Estimation of Thermonuclear Reaction Rates for Deuterium+Deuterium Reactions

NASA Astrophysics Data System (ADS)

Gómez Iñesta, Á.; Iliadis, C.; Coc, A.

2017-11-01

The study of d+d reactions is of major interest since their reaction rates affect the predicted abundances of D, 3He, and 7Li. In particular, recent measurements of primordial D/H ratios call for reduced uncertainties in the theoretical abundances predicted by Big Bang nucleosynthesis (BBN). Different authors have studied reactions involved in BBN by incorporating new experimental data and a careful treatment of systematic and probabilistic uncertainties. To analyze the experimental data, Coc et al. used results of ab initio models for the theoretical calculation of the energy dependence of S-factors in conjunction with traditional statistical methods based on χ 2 minimization. Bayesian methods have now spread to many scientific fields and provide numerous advantages in data analysis. Astrophysical S-factors and reaction rates using Bayesian statistics were calculated by Iliadis et al. Here we present a similar analysis for two d+d reactions, d(d, n)3He and d(d, p)3H, that has been translated into a total decrease of the predicted D/H value by 0.16%.

Bayesian Approach to the Joint Inversion of Gravity and Magnetic Data, with Application to the Ismenius Area of Mars

NASA Technical Reports Server (NTRS)

Jewell, Jeffrey B.; Raymond, C.; Smrekar, S.; Millbury, C.

2004-01-01

This viewgraph presentation reviews a Bayesian approach to the inversion of gravity and magnetic data with specific application to the Ismenius Area of Mars. Many inverse problems encountered in geophysics and planetary science are well known to be non-unique (i.e. inversion of gravity the density structure of a body). In hopes of reducing the non-uniqueness of solutions, there has been interest in the joint analysis of data. An example is the joint inversion of gravity and magnetic data, with the assumption that the same physical anomalies generate both the observed magnetic and gravitational anomalies. In this talk, we formulate the joint analysis of different types of data in a Bayesian framework and apply the formalism to the inference of the density and remanent magnetization structure for a local region in the Ismenius area of Mars. The Bayesian approach allows prior information or constraints in the solutions to be incorporated in the inversion, with the "best" solutions those whose forward predictions most closely match the data while remaining consistent with assumed constraints. The application of this framework to the inversion of gravity and magnetic data on Mars reveals two typical challenges - the forward predictions of the data have a linear dependence on some of the quantities of interest, and non-linear dependence on others (termed the "linear" and "non-linear" variables, respectively). For observations with Gaussian noise, a Bayesian approach to inversion for "linear" variables reduces to a linear filtering problem, with an explicitly computable "error" matrix. However, for models whose forward predictions have non-linear dependencies, inference is no longer given by such a simple linear problem, and moreover, the uncertainty in the solution is no longer completely specified by a computable "error matrix". It is therefore important to develop methods for sampling from the full Bayesian posterior to provide a complete and statistically consistent picture of model uncertainty, and what has been learned from observations. We will discuss advanced numerical techniques, including Monte Carlo Markov

Bayesian Poisson hierarchical models for crash data analysis: Investigating the impact of model choice on site-specific predictions.

PubMed

Khazraee, S Hadi; Johnson, Valen; Lord, Dominique

2018-08-01

The Poisson-gamma (PG) and Poisson-lognormal (PLN) regression models are among the most popular means for motor vehicle crash data analysis. Both models belong to the Poisson-hierarchical family of models. While numerous studies have compared the overall performance of alternative Bayesian Poisson-hierarchical models, little research has addressed the impact of model choice on the expected crash frequency prediction at individual sites. This paper sought to examine whether there are any trends among candidate models predictions e.g., that an alternative model's prediction for sites with certain conditions tends to be higher (or lower) than that from another model. In addition to the PG and PLN models, this research formulated a new member of the Poisson-hierarchical family of models: the Poisson-inverse gamma (PIGam). Three field datasets (from Texas, Michigan and Indiana) covering a wide range of over-dispersion characteristics were selected for analysis. This study demonstrated that the model choice can be critical when the calibrated models are used for prediction at new sites, especially when the data are highly over-dispersed. For all three datasets, the PIGam model would predict higher expected crash frequencies than would the PLN and PG models, in order, indicating a clear link between the models predictions and the shape of their mixing distributions (i.e., gamma, lognormal, and inverse gamma, respectively). The thicker tail of the PIGam and PLN models (in order) may provide an advantage when the data are highly over-dispersed. The analysis results also illustrated a major deficiency of the Deviance Information Criterion (DIC) in comparing the goodness-of-fit of hierarchical models; models with drastically different set of coefficients (and thus predictions for new sites) may yield similar DIC values, because the DIC only accounts for the parameters in the lowest (observation) level of the hierarchy and ignores the higher levels (regression coefficients). Copyright © 2018. Published by Elsevier Ltd.

Factors affecting GEBV accuracy with single-step Bayesian models.

PubMed

Zhou, Lei; Mrode, Raphael; Zhang, Shengli; Zhang, Qin; Li, Bugao; Liu, Jian-Feng

2018-01-01

A single-step approach to obtain genomic prediction was first proposed in 2009. Many studies have investigated the components of GEBV accuracy in genomic selection. However, it is still unclear how the population structure and the relationships between training and validation populations influence GEBV accuracy in terms of single-step analysis. Here, we explored the components of GEBV accuracy in single-step Bayesian analysis with a simulation study. Three scenarios with various numbers of QTL (5, 50, and 500) were simulated. Three models were implemented to analyze the simulated data: single-step genomic best linear unbiased prediction (GBLUP; SSGBLUP), single-step BayesA (SS-BayesA), and single-step BayesB (SS-BayesB). According to our results, GEBV accuracy was influenced by the relationships between the training and validation populations more significantly for ungenotyped animals than for genotyped animals. SS-BayesA/BayesB showed an obvious advantage over SSGBLUP with the scenarios of 5 and 50 QTL. SS-BayesB model obtained the lowest accuracy with the 500 QTL in the simulation. SS-BayesA model was the most efficient and robust considering all QTL scenarios. Generally, both the relationships between training and validation populations and LD between markers and QTL contributed to GEBV accuracy in the single-step analysis, and the advantages of single-step Bayesian models were more apparent when the trait is controlled by fewer QTL.

Uncertainty quantification for nuclear density functional theory and information content of new measurements.

PubMed

McDonnell, J D; Schunck, N; Higdon, D; Sarich, J; Wild, S M; Nazarewicz, W

2015-03-27

Statistical tools of uncertainty quantification can be used to assess the information content of measured observables with respect to present-day theoretical models, to estimate model errors and thereby improve predictive capability, to extrapolate beyond the regions reached by experiment, and to provide meaningful input to applications and planned measurements. To showcase new opportunities offered by such tools, we make a rigorous analysis of theoretical statistical uncertainties in nuclear density functional theory using Bayesian inference methods. By considering the recent mass measurements from the Canadian Penning Trap at Argonne National Laboratory, we demonstrate how the Bayesian analysis and a direct least-squares optimization, combined with high-performance computing, can be used to assess the information content of the new data with respect to a model based on the Skyrme energy density functional approach. Employing the posterior probability distribution computed with a Gaussian process emulator, we apply the Bayesian framework to propagate theoretical statistical uncertainties in predictions of nuclear masses, two-neutron dripline, and fission barriers. Overall, we find that the new mass measurements do not impose a constraint that is strong enough to lead to significant changes in the model parameters. The example discussed in this study sets the stage for quantifying and maximizing the impact of new measurements with respect to current modeling and guiding future experimental efforts, thus enhancing the experiment-theory cycle in the scientific method.

Bayesian Framework Approach for Prognostic Studies in Electrolytic Capacitor under Thermal Overstress Conditions

DTIC Science & Technology

2012-09-01

make end of life ( EOL ) and remaining useful life (RUL) estimations. Model-based prognostics approaches perform these tasks with the help of first...in parameters Degradation Modeling Parameter estimation Prediction Thermal / Electrical Stress Experimental Data State Space model RUL EOL ...distribution at given single time point kP , and use this for multi-step predictions to EOL . There are several methods which exits for selecting the sigma

A Rational Analysis of Rule-Based Concept Learning

ERIC Educational Resources Information Center

Goodman, Noah D.; Tenenbaum, Joshua B.; Feldman, Jacob; Griffiths, Thomas L.

2008-01-01

This article proposes a new model of human concept learning that provides a rational analysis of learning feature-based concepts. This model is built upon Bayesian inference for a grammatically structured hypothesis space--a concept language of logical rules. This article compares the model predictions to human generalization judgments in several…

Comparison of two statistical methods for probability prediction of monthly precipitation during summer over Huaihe River Basin in China, and applications in runoff prediction based on hydrological model

NASA Astrophysics Data System (ADS)

Liu, L.; Du, L.; Liao, Y.

2017-12-01

Based on the ensemble hindcast dataset of CSM1.1m by NCC, CMA, Bayesian merging models and a two-step statistical model are developed and employed to predict monthly grid/station precipitation in the Huaihe River China during summer at the lead-time of 1 to 3 months. The hindcast datasets span a period of 1991 to 2014. The skill of the two models is evaluated using area under the ROC curve (AUC) in a leave-one-out cross-validation framework, and is compared to the skill of CSM1.1m. CSM1.1m has highest skill for summer precipitation from April while lowest from May, and has highest skill for precipitation in June but lowest for precipitation in July. Compared with raw outputs of climate models, some schemes of the two approaches have higher skill for the prediction from March and May, but almost schemes have lower skill for prediction from April. Compared to two-step approach, one sampling scheme of Bayesian merging approach has higher skill for the prediction from March, but has lower skill from May. The results suggest that there is potential to apply the two statistical models for monthly precipitation forecast in summer from March and from May over Huaihe River basin, but is potential to apply CSM1.1m forecast from April. Finally, the summer runoff during 1991 to 2014 is simulated based on one hydrological model using the climate hindcast of CSM1.1m and the two statistical models.

Validation of the thermal challenge problem using Bayesian Belief Networks.

DOE Office of Scientific and Technical Information (OSTI.GOV)

McFarland, John; Swiler, Laura Painton

The thermal challenge problem has been developed at Sandia National Laboratories as a testbed for demonstrating various types of validation approaches and prediction methods. This report discusses one particular methodology to assess the validity of a computational model given experimental data. This methodology is based on Bayesian Belief Networks (BBNs) and can incorporate uncertainty in experimental measurements, in physical quantities, and model uncertainties. The approach uses the prior and posterior distributions of model output to compute a validation metric based on Bayesian hypothesis testing (a Bayes' factor). This report discusses various aspects of the BBN, specifically in the context ofmore » the thermal challenge problem. A BBN is developed for a given set of experimental data in a particular experimental configuration. The development of the BBN and the method for ''solving'' the BBN to develop the posterior distribution of model output through Monte Carlo Markov Chain sampling is discussed in detail. The use of the BBN to compute a Bayes' factor is demonstrated.« less

Cross-validation to select Bayesian hierarchical models in phylogenetics.

PubMed

Duchêne, Sebastián; Duchêne, David A; Di Giallonardo, Francesca; Eden, John-Sebastian; Geoghegan, Jemma L; Holt, Kathryn E; Ho, Simon Y W; Holmes, Edward C

2016-05-26

Recent developments in Bayesian phylogenetic models have increased the range of inferences that can be drawn from molecular sequence data. Accordingly, model selection has become an important component of phylogenetic analysis. Methods of model selection generally consider the likelihood of the data under the model in question. In the context of Bayesian phylogenetics, the most common approach involves estimating the marginal likelihood, which is typically done by integrating the likelihood across model parameters, weighted by the prior. Although this method is accurate, it is sensitive to the presence of improper priors. We explored an alternative approach based on cross-validation that is widely used in evolutionary analysis. This involves comparing models according to their predictive performance. We analysed simulated data and a range of viral and bacterial data sets using a cross-validation approach to compare a variety of molecular clock and demographic models. Our results show that cross-validation can be effective in distinguishing between strict- and relaxed-clock models and in identifying demographic models that allow growth in population size over time. In most of our empirical data analyses, the model selected using cross-validation was able to match that selected using marginal-likelihood estimation. The accuracy of cross-validation appears to improve with longer sequence data, particularly when distinguishing between relaxed-clock models. Cross-validation is a useful method for Bayesian phylogenetic model selection. This method can be readily implemented even when considering complex models where selecting an appropriate prior for all parameters may be difficult.

Bayesian analysis of input uncertainty in hydrological modeling: 2. Application

NASA Astrophysics Data System (ADS)

Kavetski, Dmitri; Kuczera, George; Franks, Stewart W.

2006-03-01

The Bayesian total error analysis (BATEA) methodology directly addresses both input and output errors in hydrological modeling, requiring the modeler to make explicit, rather than implicit, assumptions about the likely extent of data uncertainty. This study considers a BATEA assessment of two North American catchments: (1) French Broad River and (2) Potomac basins. It assesses the performance of the conceptual Variable Infiltration Capacity (VIC) model with and without accounting for input (precipitation) uncertainty. The results show the considerable effects of precipitation errors on the predicted hydrographs (especially the prediction limits) and on the calibrated parameters. In addition, the performance of BATEA in the presence of severe model errors is analyzed. While BATEA allows a very direct treatment of input uncertainty and yields some limited insight into model errors, it requires the specification of valid error models, which are currently poorly understood and require further work. Moreover, it leads to computationally challenging highly dimensional problems. For some types of models, including the VIC implemented using robust numerical methods, the computational cost of BATEA can be reduced using Newton-type methods.

Validation and sensitivity of the FINE Bayesian network for forecasting aquatic exposure to nano-silver.

PubMed

Money, Eric S; Barton, Lauren E; Dawson, Joseph; Reckhow, Kenneth H; Wiesner, Mark R

2014-03-01

The adaptive nature of the Forecasting the Impacts of Nanomaterials in the Environment (FINE) Bayesian network is explored. We create an updated FINE model (FINEAgNP-2) for predicting aquatic exposure concentrations of silver nanoparticles (AgNP) by combining the expert-based parameters from the baseline model established in previous work with literature data related to particle behavior, exposure, and nano-ecotoxicology via parameter learning. We validate the AgNP forecast from the updated model using mesocosm-scale field data and determine the sensitivity of several key variables to changes in environmental conditions, particle characteristics, and particle fate. Results show that the prediction accuracy of the FINEAgNP-2 model increased approximately 70% over the baseline model, with an error rate of only 20%, suggesting that FINE is a reliable tool to predict aquatic concentrations of nano-silver. Sensitivity analysis suggests that fractal dimension, particle diameter, conductivity, time, and particle fate have the most influence on aquatic exposure given the current knowledge; however, numerous knowledge gaps can be identified to suggest further research efforts that will reduce the uncertainty in subsequent exposure and risk forecasts. Copyright © 2013 Elsevier B.V. All rights reserved.

Maximum likelihood Bayesian model averaging and its predictive analysis for groundwater reactive transport models

DOE PAGES

Lu, Dan; Ye, Ming; Curtis, Gary P.

2015-08-01

While Bayesian model averaging (BMA) has been widely used in groundwater modeling, it is infrequently applied to groundwater reactive transport modeling because of multiple sources of uncertainty in the coupled hydrogeochemical processes and because of the long execution time of each model run. To resolve these problems, this study analyzed different levels of uncertainty in a hierarchical way, and used the maximum likelihood version of BMA, i.e., MLBMA, to improve the computational efficiency. Our study demonstrates the applicability of MLBMA to groundwater reactive transport modeling in a synthetic case in which twenty-seven reactive transport models were designed to predict themore » reactive transport of hexavalent uranium (U(VI)) based on observations at a former uranium mill site near Naturita, CO. Moreover, these reactive transport models contain three uncertain model components, i.e., parameterization of hydraulic conductivity, configuration of model boundary, and surface complexation reactions that simulate U(VI) adsorption. These uncertain model components were aggregated into the alternative models by integrating a hierarchical structure into MLBMA. The modeling results of the individual models and MLBMA were analyzed to investigate their predictive performance. The predictive logscore results show that MLBMA generally outperforms the best model, suggesting that using MLBMA is a sound strategy to achieve more robust model predictions relative to a single model. MLBMA works best when the alternative models are structurally distinct and have diverse model predictions. When correlation in model structure exists, two strategies were used to improve predictive performance by retaining structurally distinct models or assigning smaller prior model probabilities to correlated models. Since the synthetic models were designed using data from the Naturita site, the results of this study are expected to provide guidance for real-world modeling. Finally, limitations of applying MLBMA to the synthetic study and future real-world modeling are discussed.« less

Bayesian Recurrent Neural Network for Language Modeling.

PubMed

Chien, Jen-Tzung; Ku, Yuan-Chu

2016-02-01

A language model (LM) is calculated as the probability of a word sequence that provides the solution to word prediction for a variety of information systems. A recurrent neural network (RNN) is powerful to learn the large-span dynamics of a word sequence in the continuous space. However, the training of the RNN-LM is an ill-posed problem because of too many parameters from a large dictionary size and a high-dimensional hidden layer. This paper presents a Bayesian approach to regularize the RNN-LM and apply it for continuous speech recognition. We aim to penalize the too complicated RNN-LM by compensating for the uncertainty of the estimated model parameters, which is represented by a Gaussian prior. The objective function in a Bayesian classification network is formed as the regularized cross-entropy error function. The regularized model is constructed not only by calculating the regularized parameters according to the maximum a posteriori criterion but also by estimating the Gaussian hyperparameter by maximizing the marginal likelihood. A rapid approximation to a Hessian matrix is developed to implement the Bayesian RNN-LM (BRNN-LM) by selecting a small set of salient outer-products. The proposed BRNN-LM achieves a sparser model than the RNN-LM. Experiments on different corpora show the robustness of system performance by applying the rapid BRNN-LM under different conditions.

Relating memory to functional performance in normal aging to dementia using hierarchical Bayesian cognitive processing models.

PubMed

Shankle, William R; Pooley, James P; Steyvers, Mark; Hara, Junko; Mangrola, Tushar; Reisberg, Barry; Lee, Michael D

2013-01-01

Determining how cognition affects functional abilities is important in Alzheimer disease and related disorders. A total of 280 patients (normal or Alzheimer disease and related disorders) received a total of 1514 assessments using the functional assessment staging test (FAST) procedure and the MCI Screen. A hierarchical Bayesian cognitive processing model was created by embedding a signal detection theory model of the MCI Screen-delayed recognition memory task into a hierarchical Bayesian framework. The signal detection theory model used latent parameters of discriminability (memory process) and response bias (executive function) to predict, simultaneously, recognition memory performance for each patient and each FAST severity group. The observed recognition memory data did not distinguish the 6 FAST severity stages, but the latent parameters completely separated them. The latent parameters were also used successfully to transform the ordinal FAST measure into a continuous measure reflecting the underlying continuum of functional severity. Hierarchical Bayesian cognitive processing models applied to recognition memory data from clinical practice settings accurately translated a latent measure of cognition into a continuous measure of functional severity for both individuals and FAST groups. Such a translation links 2 levels of brain information processing and may enable more accurate correlations with other levels, such as those characterized by biomarkers.

A data-driven SVR model for long-term runoff prediction and uncertainty analysis based on the Bayesian framework

NASA Astrophysics Data System (ADS)

Liang, Zhongmin; Li, Yujie; Hu, Yiming; Li, Binquan; Wang, Jun

2017-06-01

Accurate and reliable long-term forecasting plays an important role in water resources management and utilization. In this paper, a hybrid model called SVR-HUP is presented to predict long-term runoff and quantify the prediction uncertainty. The model is created based on three steps. First, appropriate predictors are selected according to the correlations between meteorological factors and runoff. Second, a support vector regression (SVR) model is structured and optimized based on the LibSVM toolbox and a genetic algorithm. Finally, using forecasted and observed runoff, a hydrologic uncertainty processor (HUP) based on a Bayesian framework is used to estimate the posterior probability distribution of the simulated values, and the associated uncertainty of prediction was quantitatively analyzed. Six precision evaluation indexes, including the correlation coefficient (CC), relative root mean square error (RRMSE), relative error (RE), mean absolute percentage error (MAPE), Nash-Sutcliffe efficiency (NSE), and qualification rate (QR), are used to measure the prediction accuracy. As a case study, the proposed approach is applied in the Han River basin, South Central China. Three types of SVR models are established to forecast the monthly, flood season and annual runoff volumes. The results indicate that SVR yields satisfactory accuracy and reliability at all three scales. In addition, the results suggest that the HUP cannot only quantify the uncertainty of prediction based on a confidence interval but also provide a more accurate single value prediction than the initial SVR forecasting result. Thus, the SVR-HUP model provides an alternative method for long-term runoff forecasting.

Dysphagia Management and Research in an Acute-Care Military Treatment Facility: The Role of Applied Informatics.

PubMed

Solomon, Nancy Pearl; Dietsch, Angela M; Dietrich-Burns, Katie E; Styrmisdottir, Edda L; Armao, Christopher S

2016-05-01

This report describes the development and preliminary analysis of a database for traumatically injured military service members with dysphagia. A multidimensional database was developed to capture clinical variables related to swallowing. Data were derived from clinical records and instrumental swallow studies, and ranged from demographics, injury characteristics, swallowing biomechanics, medications, and standardized tools (e.g., Glasgow Coma Scale, Penetration-Aspiration Scale). Bayesian Belief Network modeling was used to analyze the data at intermediate points, guide data collection, and predict outcomes. Predictive models were validated with independent data via receiver operating characteristic curves. The first iteration of the model (n = 48) revealed variables that could be collapsed for the second model (n = 96). The ability to predict recovery from dysphagia improved from the second to third models (area under the curve = 0.68 to 0.86). The third model, based on 161 cases, revealed "initial diet restrictions" as first-degree, and "Glasgow Coma Scale, intubation history, and diet change" as second-degree associates for diet restrictions at discharge. This project demonstrates the potential for bioinformatics to advance understanding of dysphagia. This database in concert with Bayesian Belief Network modeling makes it possible to explore predictive relationships between injuries and swallowing function, individual variability in recovery, and appropriate treatment options. Reprint & Copyright © 2016 Association of Military Surgeons of the U.S.

Ant colony optimization algorithm for interpretable Bayesian classifiers combination: application to medical predictions.

PubMed

Bouktif, Salah; Hanna, Eileen Marie; Zaki, Nazar; Abu Khousa, Eman

2014-01-01

Prediction and classification techniques have been well studied by machine learning researchers and developed for several real-word problems. However, the level of acceptance and success of prediction models are still below expectation due to some difficulties such as the low performance of prediction models when they are applied in different environments. Such a problem has been addressed by many researchers, mainly from the machine learning community. A second problem, principally raised by model users in different communities, such as managers, economists, engineers, biologists, and medical practitioners, etc., is the prediction models' interpretability. The latter is the ability of a model to explain its predictions and exhibit the causality relationships between the inputs and the outputs. In the case of classification, a successful way to alleviate the low performance is to use ensemble classiers. It is an intuitive strategy to activate collaboration between different classifiers towards a better performance than individual classier. Unfortunately, ensemble classifiers method do not take into account the interpretability of the final classification outcome. It even worsens the original interpretability of the individual classifiers. In this paper we propose a novel implementation of classifiers combination approach that does not only promote the overall performance but also preserves the interpretability of the resulting model. We propose a solution based on Ant Colony Optimization and tailored for the case of Bayesian classifiers. We validate our proposed solution with case studies from medical domain namely, heart disease and Cardiotography-based predictions, problems where interpretability is critical to make appropriate clinical decisions. The datasets, Prediction Models and software tool together with supplementary materials are available at http://faculty.uaeu.ac.ae/salahb/ACO4BC.htm.

Characterization of a Saccharomyces cerevisiae fermentation process for production of a therapeutic recombinant protein using a multivariate Bayesian approach.

PubMed

Fu, Zhibiao; Baker, Daniel; Cheng, Aili; Leighton, Julie; Appelbaum, Edward; Aon, Juan

2016-05-01

The principle of quality by design (QbD) has been widely applied to biopharmaceutical manufacturing processes. Process characterization is an essential step to implement the QbD concept to establish the design space and to define the proven acceptable ranges (PAR) for critical process parameters (CPPs). In this study, we present characterization of a Saccharomyces cerevisiae fermentation process using risk assessment analysis, statistical design of experiments (DoE), and the multivariate Bayesian predictive approach. The critical quality attributes (CQAs) and CPPs were identified with a risk assessment. The statistical model for each attribute was established using the results from the DoE study with consideration given to interactions between CPPs. Both the conventional overlapping contour plot and the multivariate Bayesian predictive approaches were used to establish the region of process operating conditions where all attributes met their specifications simultaneously. The quantitative Bayesian predictive approach was chosen to define the PARs for the CPPs, which apply to the manufacturing control strategy. Experience from the 10,000 L manufacturing scale process validation, including 64 continued process verification batches, indicates that the CPPs remain under a state of control and within the established PARs. The end product quality attributes were within their drug substance specifications. The probability generated with the Bayesian approach was also used as a tool to assess CPP deviations. This approach can be extended to develop other production process characterization and quantify a reliable operating region. © 2016 American Institute of Chemical Engineers Biotechnol. Prog., 32:799-812, 2016. © 2016 American Institute of Chemical Engineers.

Population pharmacokinetics and maximum a posteriori probability Bayesian estimator of abacavir: application of individualized therapy in HIV-infected infants and toddlers.

PubMed

Zhao, Wei; Cella, Massimo; Della Pasqua, Oscar; Burger, David; Jacqz-Aigrain, Evelyne

2012-04-01

Abacavir is used to treat HIV infection in both adults and children. The recommended paediatric dose is 8 mg kg(-1) twice daily up to a maximum of 300 mg twice daily. Weight was identified as the central covariate influencing pharmacokinetics of abacavir in children. A population pharmacokinetic model was developed to describe both once and twice daily pharmacokinetic profiles of abacavir in infants and toddlers. Standard dosage regimen is associated with large interindividual variability in abacavir concentrations. A maximum a posteriori probability Bayesian estimator of AUC(0-) (t) based on three time points (0, 1 or 2, and 3 h) is proposed to support area under the concentration-time curve (AUC) targeted individualized therapy in infants and toddlers. To develop a population pharmacokinetic model for abacavir in HIV-infected infants and toddlers, which will be used to describe both once and twice daily pharmacokinetic profiles, identify covariates that explain variability and propose optimal time points to optimize the area under the concentration-time curve (AUC) targeted dosage and individualize therapy. The pharmacokinetics of abacavir was described with plasma concentrations from 23 patients using nonlinear mixed-effects modelling (NONMEM) software. A two-compartment model with first-order absorption and elimination was developed. The final model was validated using bootstrap, visual predictive check and normalized prediction distribution errors. The Bayesian estimator was validated using the cross-validation and simulation-estimation method. The typical population pharmacokinetic parameters and relative standard errors (RSE) were apparent systemic clearance (CL) 13.4 () h−1 (RSE 6.3%), apparent central volume of distribution 4.94 () (RSE 28.7%), apparent peripheral volume of distribution 8.12 () (RSE14.2%), apparent intercompartment clearance 1.25 () h−1 (RSE 16.9%) and absorption rate constant 0.758 h−1 (RSE 5.8%). The covariate analysis identified weight as the individual factor influencing the apparent oral clearance: CL = 13.4 × (weight/12)1.14. The maximum a posteriori probability Bayesian estimator, based on three concentrations measured at 0, 1 or 2, and 3 h after drug intake allowed predicting individual AUC0–t. The population pharmacokinetic model developed for abacavir in HIV-infected infants and toddlers accurately described both once and twice daily pharmacokinetic profiles. The maximum a posteriori probability Bayesian estimator of AUC(0-) (t) was developed from the final model and can be used routinely to optimize individual dosing. © 2011 The Authors. British Journal of Clinical Pharmacology © 2011 The British Pharmacological Society.

Hierarchical Bayesian Logistic Regression to forecast metabolic control in type 2 DM patients.

PubMed

Dagliati, Arianna; Malovini, Alberto; Decata, Pasquale; Cogni, Giulia; Teliti, Marsida; Sacchi, Lucia; Cerra, Carlo; Chiovato, Luca; Bellazzi, Riccardo

2016-01-01

In this work we present our efforts in building a model able to forecast patients' changes in clinical conditions when repeated measurements are available. In this case the available risk calculators are typically not applicable. We propose a Hierarchical Bayesian Logistic Regression model, which allows taking into account individual and population variability in model parameters estimate. The model is used to predict metabolic control and its variation in type 2 diabetes mellitus. In particular we have analyzed a population of more than 1000 Italian type 2 diabetic patients, collected within the European project Mosaic. The results obtained in terms of Matthews Correlation Coefficient are significantly better than the ones gathered with standard logistic regression model, based on data pooling.

A spectral-spatial-dynamic hierarchical Bayesian (SSD-HB) model for estimating soybean yield

NASA Astrophysics Data System (ADS)

Kazama, Yoriko; Kujirai, Toshihiro

2014-10-01

A method called a "spectral-spatial-dynamic hierarchical-Bayesian (SSD-HB) model," which can deal with many parameters (such as spectral and weather information all together) by reducing the occurrence of multicollinearity, is proposed. Experiments conducted on soybean yields in Brazil fields with a RapidEye satellite image indicate that the proposed SSD-HB model can predict soybean yield with a higher degree of accuracy than other estimation methods commonly used in remote-sensing applications. In the case of the SSD-HB model, the mean absolute error between estimated yield of the target area and actual yield is 0.28 t/ha, compared to 0.34 t/ha when conventional PLS regression was applied, showing the potential effectiveness of the proposed model.

Intelligent Decisions Need Intelligent Choice of Models and Data - a Bayesian Justifiability Analysis for Models with Vastly Different Complexity

NASA Astrophysics Data System (ADS)

Nowak, W.; Schöniger, A.; Wöhling, T.; Illman, W. A.

2016-12-01

Model-based decision support requires justifiable models with good predictive capabilities. This, in turn, calls for a fine adjustment between predictive accuracy (small systematic model bias that can be achieved with rather complex models), and predictive precision (small predictive uncertainties that can be achieved with simpler models with fewer parameters). The implied complexity/simplicity trade-off depends on the availability of informative data for calibration. If not available, additional data collection can be planned through optimal experimental design. We present a model justifiability analysis that can compare models of vastly different complexity. It rests on Bayesian model averaging (BMA) to investigate the complexity/performance trade-off dependent on data availability. Then, we disentangle the complexity component from the performance component. We achieve this by replacing actually observed data by realizations of synthetic data predicted by the models. This results in a "model confusion matrix". Based on this matrix, the modeler can identify the maximum model complexity that can be justified by the available (or planned) amount and type of data. As a side product, the matrix quantifies model (dis-)similarity. We apply this analysis to aquifer characterization via hydraulic tomography, comparing four models with a vastly different number of parameters (from a homogeneous model to geostatistical random fields). As a testing scenario, we consider hydraulic tomography data. Using subsets of these data, we determine model justifiability as a function of data set size. The test case shows that geostatistical parameterization requires a substantial amount of hydraulic tomography data to be justified, while a zonation-based model can be justified with more limited data set sizes. The actual model performance (as opposed to model justifiability), however, depends strongly on the quality of prior geological information.

Bayesian model selection validates a biokinetic model for zirconium processing in humans

PubMed Central

2012-01-01

Background In radiation protection, biokinetic models for zirconium processing are of crucial importance in dose estimation and further risk analysis for humans exposed to this radioactive substance. They provide limiting values of detrimental effects and build the basis for applications in internal dosimetry, the prediction for radioactive zirconium retention in various organs as well as retrospective dosimetry. Multi-compartmental models are the tool of choice for simulating the processing of zirconium. Although easily interpretable, determining the exact compartment structure and interaction mechanisms is generally daunting. In the context of observing the dynamics of multiple compartments, Bayesian methods provide efficient tools for model inference and selection. Results We are the first to apply a Markov chain Monte Carlo approach to compute Bayes factors for the evaluation of two competing models for zirconium processing in the human body after ingestion. Based on in vivo measurements of human plasma and urine levels we were able to show that a recently published model is superior to the standard model of the International Commission on Radiological Protection. The Bayes factors were estimated by means of the numerically stable thermodynamic integration in combination with a recently developed copula-based Metropolis-Hastings sampler. Conclusions In contrast to the standard model the novel model predicts lower accretion of zirconium in bones. This results in lower levels of noxious doses for exposed individuals. Moreover, the Bayesian approach allows for retrospective dose assessment, including credible intervals for the initially ingested zirconium, in a significantly more reliable fashion than previously possible. All methods presented here are readily applicable to many modeling tasks in systems biology. PMID:22863152

Bayesian spatio-temporal discard model in a demersal trawl fishery

NASA Astrophysics Data System (ADS)

Grazia Pennino, M.; Muñoz, Facundo; Conesa, David; López-Quílez, Antonio; Bellido, José M.

2014-07-01

Spatial management of discards has recently been proposed as a useful tool for the protection of juveniles, by reducing discard rates and can be used as a buffer against management errors and recruitment failure. In this study Bayesian hierarchical spatial models have been used to analyze about 440 trawl fishing operations of two different metiers, sampled between 2009 and 2012, in order to improve our understanding of factors that influence the quantity of discards and to identify their spatio-temporal distribution in the study area. Our analysis showed that the relative importance of each variable was different for each metier, with a few similarities. In particular, the random vessel effect and seasonal variability were identified as main driving variables for both metiers. Predictive maps of the abundance of discards and maps of the posterior mean of the spatial component show several hot spots with high discard concentration for each metier. We argue how the seasonal/spatial effects, and the knowledge about the factors influential to discarding, could potentially be exploited as potential mitigation measures for future fisheries management strategies. However, misidentification of hotspots and uncertain predictions can culminate in inappropriate mitigation practices which can sometimes be irreversible. The proposed Bayesian spatial method overcomes these issues, since it offers a unified approach which allows the incorporation of spatial random-effect terms, spatial correlation of the variables and the uncertainty of the parameters in the modeling process, resulting in a better quantification of the uncertainty and accurate predictions.

Spatiotemporal Modeling of Ozone Levels in Quebec (Canada): A Comparison of Kriging, Land-Use Regression (LUR), and Combined Bayesian Maximum Entropy–LUR Approaches

PubMed Central

Adam-Poupart, Ariane; Brand, Allan; Fournier, Michel; Jerrett, Michael

2014-01-01

Background: Ambient air ozone (O3) is a pulmonary irritant that has been associated with respiratory health effects including increased lung inflammation and permeability, airway hyperreactivity, respiratory symptoms, and decreased lung function. Estimation of O3 exposure is a complex task because the pollutant exhibits complex spatiotemporal patterns. To refine the quality of exposure estimation, various spatiotemporal methods have been developed worldwide. Objectives: We sought to compare the accuracy of three spatiotemporal models to predict summer ground-level O3 in Quebec, Canada. Methods: We developed a land-use mixed-effects regression (LUR) model based on readily available data (air quality and meteorological monitoring data, road networks information, latitude), a Bayesian maximum entropy (BME) model incorporating both O3 monitoring station data and the land-use mixed model outputs (BME-LUR), and a kriging method model based only on available O3 monitoring station data (BME kriging). We performed leave-one-station-out cross-validation and visually assessed the predictive capability of each model by examining the mean temporal and spatial distributions of the average estimated errors. Results: The BME-LUR was the best predictive model (R2 = 0.653) with the lowest root mean-square error (RMSE ;7.06 ppb), followed by the LUR model (R2 = 0.466, RMSE = 8.747) and the BME kriging model (R2 = 0.414, RMSE = 9.164). Conclusions: Our findings suggest that errors of estimation in the interpolation of O3 concentrations with BME can be greatly reduced by incorporating outputs from a LUR model developed with readily available data. Citation: Adam-Poupart A, Brand A, Fournier M, Jerrett M, Smargiassi A. 2014. Spatiotemporal modeling of ozone levels in Quebec (Canada): a comparison of kriging, land-use regression (LUR), and combined Bayesian maximum entropy–LUR approaches. Environ Health Perspect 122:970–976; http://dx.doi.org/10.1289/ehp.1306566 PMID:24879650

A Bayesian Poisson-lognormal Model for Count Data for Multiple-Trait Multiple-Environment Genomic-Enabled Prediction.

PubMed

Montesinos-López, Osval A; Montesinos-López, Abelardo; Crossa, José; Toledo, Fernando H; Montesinos-López, José C; Singh, Pawan; Juliana, Philomin; Salinas-Ruiz, Josafhat

2017-05-05

When a plant scientist wishes to make genomic-enabled predictions of multiple traits measured in multiple individuals in multiple environments, the most common strategy for performing the analysis is to use a single trait at a time taking into account genotype × environment interaction (G × E), because there is a lack of comprehensive models that simultaneously take into account the correlated counting traits and G × E. For this reason, in this study we propose a multiple-trait and multiple-environment model for count data. The proposed model was developed under the Bayesian paradigm for which we developed a Markov Chain Monte Carlo (MCMC) with noninformative priors. This allows obtaining all required full conditional distributions of the parameters leading to an exact Gibbs sampler for the posterior distribution. Our model was tested with simulated data and a real data set. Results show that the proposed multi-trait, multi-environment model is an attractive alternative for modeling multiple count traits measured in multiple environments. Copyright © 2017 Montesinos-López et al.

Laminar fMRI and computational theories of brain function.

PubMed

Stephan, K E; Petzschner, F H; Kasper, L; Bayer, J; Wellstein, K V; Stefanics, G; Pruessmann, K P; Heinzle, J

2017-11-02

Recently developed methods for functional MRI at the resolution of cortical layers (laminar fMRI) offer a novel window into neurophysiological mechanisms of cortical activity. Beyond physiology, laminar fMRI also offers an unprecedented opportunity to test influential theories of brain function. Specifically, hierarchical Bayesian theories of brain function, such as predictive coding, assign specific computational roles to different cortical layers. Combined with computational models, laminar fMRI offers a unique opportunity to test these proposals noninvasively in humans. This review provides a brief overview of predictive coding and related hierarchical Bayesian theories, summarises their predictions with regard to layered cortical computations, examines how these predictions could be tested by laminar fMRI, and considers methodological challenges. We conclude by discussing the potential of laminar fMRI for clinically useful computational assays of layer-specific information processing. Copyright © 2017 Elsevier Inc. All rights reserved.

Propagation of Bayesian Belief for Near-Real Time Statistical Assessment of Geosynchronous Satellite Status Based on Non-Resolved Photometry Data

DTIC Science & Technology

2014-09-01

of the BRDF for the Body and Panel. In order to provide a continuously updated baseline, the Photometry Model application is performed using a...brightness to its predicted brightness. The brightness predictions can be obtained using any analytical model chosen by the user. The inference for a...the analytical model as possible; and to mitigate the effect of bias that could be introduced by the choice of analytical model . It considers that a

Bayesian whole-genome prediction and genome-wide association analysis with missing genotypes using variable selection

USDA-ARS?s Scientific Manuscript database

Single-step Genomic Best Linear Unbiased Predictor (ssGBLUP) has become increasingly popular for whole-genome prediction (WGP) modeling as it utilizes any available pedigree and phenotypes on both genotyped and non-genotyped individuals. The WGP accuracy of ssGBLUP has been demonstrated to be greate...

Bayesian structural equation modeling: a more flexible representation of substantive theory.

PubMed

Muthén, Bengt; Asparouhov, Tihomir

2012-09-01

This article proposes a new approach to factor analysis and structural equation modeling using Bayesian analysis. The new approach replaces parameter specifications of exact zeros with approximate zeros based on informative, small-variance priors. It is argued that this produces an analysis that better reflects substantive theories. The proposed Bayesian approach is particularly beneficial in applications where parameters are added to a conventional model such that a nonidentified model is obtained if maximum-likelihood estimation is applied. This approach is useful for measurement aspects of latent variable modeling, such as with confirmatory factor analysis, and the measurement part of structural equation modeling. Two application areas are studied, cross-loadings and residual correlations in confirmatory factor analysis. An example using a full structural equation model is also presented, showing an efficient way to find model misspecification. The approach encompasses 3 elements: model testing using posterior predictive checking, model estimation, and model modification. Monte Carlo simulations and real data are analyzed using Mplus. The real-data analyses use data from Holzinger and Swineford's (1939) classic mental abilities study, Big Five personality factor data from a British survey, and science achievement data from the National Educational Longitudinal Study of 1988.

Comparing flood loss models of different complexity

NASA Astrophysics Data System (ADS)

Schröter, Kai; Kreibich, Heidi; Vogel, Kristin; Riggelsen, Carsten; Scherbaum, Frank; Merz, Bruno

2013-04-01

Any deliberation on flood risk requires the consideration of potential flood losses. In particular, reliable flood loss models are needed to evaluate cost-effectiveness of mitigation measures, to assess vulnerability, for comparative risk analysis and financial appraisal during and after floods. In recent years, considerable improvements have been made both concerning the data basis and the methodological approaches used for the development of flood loss models. Despite of that, flood loss models remain an important source of uncertainty. Likewise the temporal and spatial transferability of flood loss models is still limited. This contribution investigates the predictive capability of different flood loss models in a split sample cross regional validation approach. For this purpose, flood loss models of different complexity, i.e. based on different numbers of explaining variables, are learned from a set of damage records that was obtained from a survey after the Elbe flood in 2002. The validation of model predictions is carried out for different flood events in the Elbe and Danube river basins in 2002, 2005 and 2006 for which damage records are available from surveys after the flood events. The models investigated are a stage-damage model, the rule based model FLEMOps+r as well as novel model approaches which are derived using data mining techniques of regression trees and Bayesian networks. The Bayesian network approach to flood loss modelling provides attractive additional information concerning the probability distribution of both model predictions and explaining variables.

Predicting the geographical distribution of two invasive termite species from occurrence data.

PubMed

Tonini, Francesco; Divino, Fabio; Lasinio, Giovanna Jona; Hochmair, Hartwig H; Scheffrahn, Rudolf H

2014-10-01

Predicting the potential habitat of species under both current and future climate change scenarios is crucial for monitoring invasive species and understanding a species' response to different environmental conditions. Frequently, the only data available on a species is the location of its occurrence (presence-only data). Using occurrence records only, two models were used to predict the geographical distribution of two destructive invasive termite species, Coptotermes gestroi (Wasmann) and Coptotermes formosanus Shiraki. The first model uses a Bayesian linear logistic regression approach adjusted for presence-only data while the second one is the widely used maximum entropy approach (Maxent). Results show that the predicted distributions of both C. gestroi and C. formosanus are strongly linked to urban development. The impact of future scenarios such as climate warming and population growth on the biotic distribution of both termite species was also assessed. Future climate warming seems to affect their projected probability of presence to a lesser extent than population growth. The Bayesian logistic approach outperformed Maxent consistently in all models according to evaluation criteria such as model sensitivity and ecological realism. The importance of further studies for an explicit treatment of residual spatial autocorrelation and a more comprehensive comparison between both statistical approaches is suggested.

Optimization of nonlinear, non-Gaussian Bayesian filtering for diagnosis and prognosis of monotonic degradation processes

NASA Astrophysics Data System (ADS)

Corbetta, Matteo; Sbarufatti, Claudio; Giglio, Marco; Todd, Michael D.

2018-05-01

The present work critically analyzes the probabilistic definition of dynamic state-space models subject to Bayesian filters used for monitoring and predicting monotonic degradation processes. The study focuses on the selection of the random process, often called process noise, which is a key perturbation source in the evolution equation of particle filtering. Despite the large number of applications of particle filtering predicting structural degradation, the adequacy of the picked process noise has not been investigated. This paper reviews existing process noise models that are typically embedded in particle filters dedicated to monitoring and predicting structural damage caused by fatigue, which is monotonic in nature. The analysis emphasizes that existing formulations of the process noise can jeopardize the performance of the filter in terms of state estimation and remaining life prediction (i.e., damage prognosis). This paper subsequently proposes an optimal and unbiased process noise model and a list of requirements that the stochastic model must satisfy to guarantee high prognostic performance. These requirements are useful for future and further implementations of particle filtering for monotonic system dynamics. The validity of the new process noise formulation is assessed against experimental fatigue crack growth data from a full-scale aeronautical structure using dedicated performance metrics.

Iowa radon leukaemia study: a hierarchical population risk model for spatially correlated exposure measured with error.

PubMed

Smith, Brian J; Zhang, Lixun; Field, R William

2007-11-10

This paper presents a Bayesian model that allows for the joint prediction of county-average radon levels and estimation of the associated leukaemia risk. The methods are motivated by radon data from an epidemiologic study of residential radon in Iowa that include 2726 outdoor and indoor measurements. Prediction of county-average radon is based on a geostatistical model for the radon data which assumes an underlying continuous spatial process. In the radon model, we account for uncertainties due to incomplete spatial coverage, spatial variability, characteristic differences between homes, and detector measurement error. The predicted radon averages are, in turn, included as a covariate in Poisson models for incident cases of acute lymphocytic (ALL), acute myelogenous (AML), chronic lymphocytic (CLL), and chronic myelogenous (CML) leukaemias reported to the Iowa cancer registry from 1973 to 2002. Since radon and leukaemia risk are modelled simultaneously in our approach, the resulting risk estimates accurately reflect uncertainties in the predicted radon exposure covariate. Posterior mean (95 per cent Bayesian credible interval) estimates of the relative risk associated with a 1 pCi/L increase in radon for ALL, AML, CLL, and CML are 0.91 (0.78-1.03), 1.01 (0.92-1.12), 1.06 (0.96-1.16), and 1.12 (0.98-1.27), respectively. Copyright 2007 John Wiley & Sons, Ltd.

How Recent History Affects Perception: The Normative Approach and Its Heuristic Approximation

PubMed Central

Raviv, Ofri; Ahissar, Merav; Loewenstein, Yonatan

2012-01-01

There is accumulating evidence that prior knowledge about expectations plays an important role in perception. The Bayesian framework is the standard computational approach to explain how prior knowledge about the distribution of expected stimuli is incorporated with noisy observations in order to improve performance. However, it is unclear what information about the prior distribution is acquired by the perceptual system over short periods of time and how this information is utilized in the process of perceptual decision making. Here we address this question using a simple two-tone discrimination task. We find that the “contraction bias”, in which small magnitudes are overestimated and large magnitudes are underestimated, dominates the pattern of responses of human participants. This contraction bias is consistent with the Bayesian hypothesis in which the true prior information is available to the decision-maker. However, a trial-by-trial analysis of the pattern of responses reveals that the contribution of most recent trials to performance is overweighted compared with the predictions of a standard Bayesian model. Moreover, we study participants' performance in a-typical distributions of stimuli and demonstrate substantial deviations from the ideal Bayesian detector, suggesting that the brain utilizes a heuristic approximation of the Bayesian inference. We propose a biologically plausible model, in which decision in the two-tone discrimination task is based on a comparison between the second tone and an exponentially-decaying average of the first tone and past tones. We show that this model accounts for both the contraction bias and the deviations from the ideal Bayesian detector hypothesis. These findings demonstrate the power of Bayesian-like heuristics in the brain, as well as their limitations in their failure to fully adapt to novel environments. PMID:23133343

Identification of transmissivity fields using a Bayesian strategy and perturbative approach

NASA Astrophysics Data System (ADS)

Zanini, Andrea; Tanda, Maria Giovanna; Woodbury, Allan D.

2017-10-01

The paper deals with the crucial problem of the groundwater parameter estimation that is the basis for efficient modeling and reclamation activities. A hierarchical Bayesian approach is developed: it uses the Akaike's Bayesian Information Criteria in order to estimate the hyperparameters (related to the covariance model chosen) and to quantify the unknown noise variance. The transmissivity identification proceeds in two steps: the first, called empirical Bayesian interpolation, uses Y* (Y = lnT) observations to interpolate Y values on a specified grid; the second, called empirical Bayesian update, improve the previous Y estimate through the addition of hydraulic head observations. The relationship between the head and the lnT has been linearized through a perturbative solution of the flow equation. In order to test the proposed approach, synthetic aquifers from literature have been considered. The aquifers in question contain a variety of boundary conditions (both Dirichelet and Neuman type) and scales of heterogeneities (σY2 = 1.0 and σY2 = 5.3). The estimated transmissivity fields were compared to the true one. The joint use of Y* and head measurements improves the estimation of Y considering both degrees of heterogeneity. Even if the variance of the strong transmissivity field can be considered high for the application of the perturbative approach, the results show the same order of approximation of the non-linear methods proposed in literature. The procedure allows to compute the posterior probability distribution of the target quantities and to quantify the uncertainty in the model prediction. Bayesian updating has advantages related both to the Monte-Carlo (MC) and non-MC approaches. In fact, as the MC methods, Bayesian updating allows computing the direct posterior probability distribution of the target quantities and as non-MC methods it has computational times in the order of seconds.

Bayesian inference for the spatio-temporal invasion of alien species.

PubMed

Cook, Alex; Marion, Glenn; Butler, Adam; Gibson, Gavin

2007-08-01

In this paper we develop a Bayesian approach to parameter estimation in a stochastic spatio-temporal model of the spread of invasive species across a landscape. To date, statistical techniques, such as logistic and autologistic regression, have outstripped stochastic spatio-temporal models in their ability to handle large numbers of covariates. Here we seek to address this problem by making use of a range of covariates describing the bio-geographical features of the landscape. Relative to regression techniques, stochastic spatio-temporal models are more transparent in their representation of biological processes. They also explicitly model temporal change, and therefore do not require the assumption that the species' distribution (or other spatial pattern) has already reached equilibrium as is often the case with standard statistical approaches. In order to illustrate the use of such techniques we apply them to the analysis of data detailing the spread of an invasive plant, Heracleum mantegazzianum, across Britain in the 20th Century using geo-referenced covariate information describing local temperature, elevation and habitat type. The use of Markov chain Monte Carlo sampling within a Bayesian framework facilitates statistical assessments of differences in the suitability of different habitat classes for H. mantegazzianum, and enables predictions of future spread to account for parametric uncertainty and system variability. Our results show that ignoring such covariate information may lead to biased estimates of key processes and implausible predictions of future distributions.

Validating neural-network refinements of nuclear mass models

NASA Astrophysics Data System (ADS)

Utama, R.; Piekarewicz, J.

2018-01-01

Background: Nuclear astrophysics centers on the role of nuclear physics in the cosmos. In particular, nuclear masses at the limits of stability are critical in the development of stellar structure and the origin of the elements. Purpose: We aim to test and validate the predictions of recently refined nuclear mass models against the newly published AME2016 compilation. Methods: The basic paradigm underlining the recently refined nuclear mass models is based on existing state-of-the-art models that are subsequently refined through the training of an artificial neural network. Bayesian inference is used to determine the parameters of the neural network so that statistical uncertainties are provided for all model predictions. Results: We observe a significant improvement in the Bayesian neural network (BNN) predictions relative to the corresponding "bare" models when compared to the nearly 50 new masses reported in the AME2016 compilation. Further, AME2016 estimates for the handful of impactful isotopes in the determination of r -process abundances are found to be in fairly good agreement with our theoretical predictions. Indeed, the BNN-improved Duflo-Zuker model predicts a root-mean-square deviation relative to experiment of σrms≃400 keV. Conclusions: Given the excellent performance of the BNN refinement in confronting the recently published AME2016 compilation, we are confident of its critical role in our quest for mass models of the highest quality. Moreover, as uncertainty quantification is at the core of the BNN approach, the improved mass models are in a unique position to identify those nuclei that will have the strongest impact in resolving some of the outstanding questions in nuclear astrophysics.

New insights into faster computation of uncertainties

NASA Astrophysics Data System (ADS)

Bhattacharya, Atreyee

2012-11-01

Heavy computation power, lengthy simulations, and an exhaustive number of model runs—often these seem like the only statistical tools that scientists have at their disposal when computing uncertainties associated with predictions, particularly in cases of environmental processes such as groundwater movement. However, calculation of uncertainties need not be as lengthy, a new study shows. Comparing two approaches—the classical Bayesian “credible interval” and a less commonly used regression-based “confidence interval” method—Lu et al. show that for many practical purposes both methods provide similar estimates of uncertainties. The advantage of the regression method is that it demands 10-1000 model runs, whereas the classical Bayesian approach requires 10,000 to millions of model runs.

Interoceptive predictions in the brain

PubMed Central

Barrett, Lisa Feldman; Simmons, W. Kyle

2016-01-01

Intuition suggests that perception follows sensation and therefore bodily feelings originate in the body. However, recent evidence goes against this logic: interoceptive experience may largely reflect limbic predictions about the expected state of the body that are constrained by ascending visceral sensations. In this Opinion article, we introduce the Embodied Predictive Interoception Coding model, which integrates an anatomical model of corticocortical connections with Bayesian active inference principles, to propose that agranular visceromotor cortices contribute to interoception by issuing interoceptive predictions. We then discuss how disruptions in interoceptive predictions could function as a common vulnerability for mental and physical illness. PMID:26016744

Genomic selection and complex trait prediction using a fast EM algorithm applied to genome-wide markers

PubMed Central

2010-01-01

Background The information provided by dense genome-wide markers using high throughput technology is of considerable potential in human disease studies and livestock breeding programs. Genome-wide association studies relate individual single nucleotide polymorphisms (SNP) from dense SNP panels to individual measurements of complex traits, with the underlying assumption being that any association is caused by linkage disequilibrium (LD) between SNP and quantitative trait loci (QTL) affecting the trait. Often SNP are in genomic regions of no trait variation. Whole genome Bayesian models are an effective way of incorporating this and other important prior information into modelling. However a full Bayesian analysis is often not feasible due to the large computational time involved. Results This article proposes an expectation-maximization (EM) algorithm called emBayesB which allows only a proportion of SNP to be in LD with QTL and incorporates prior information about the distribution of SNP effects. The posterior probability of being in LD with at least one QTL is calculated for each SNP along with estimates of the hyperparameters for the mixture prior. A simulated example of genomic selection from an international workshop is used to demonstrate the features of the EM algorithm. The accuracy of prediction is comparable to a full Bayesian analysis but the EM algorithm is considerably faster. The EM algorithm was accurate in locating QTL which explained more than 1% of the total genetic variation. A computational algorithm for very large SNP panels is described. Conclusions emBayesB is a fast and accurate EM algorithm for implementing genomic selection and predicting complex traits by mapping QTL in genome-wide dense SNP marker data. Its accuracy is similar to Bayesian methods but it takes only a fraction of the time. PMID:20969788

Can Bayesian Theories of Autism Spectrum Disorder Help Improve Clinical Practice?

PubMed

Haker, Helene; Schneebeli, Maya; Stephan, Klaas Enno

2016-01-01

Diagnosis and individualized treatment of autism spectrum disorder (ASD) represent major problems for contemporary psychiatry. Tackling these problems requires guidance by a pathophysiological theory. In this paper, we consider recent theories that re-conceptualize ASD from a "Bayesian brain" perspective, which posit that the core abnormality of ASD resides in perceptual aberrations due to a disbalance in the precision of prediction errors (sensory noise) relative to the precision of predictions (prior beliefs). This results in percepts that are dominated by sensory inputs and less guided by top-down regularization and shifts the perceptual focus to detailed aspects of the environment with difficulties in extracting meaning. While these Bayesian theories have inspired ongoing empirical studies, their clinical implications have not yet been carved out. Here, we consider how this Bayesian perspective on disease mechanisms in ASD might contribute to improving clinical care for affected individuals. Specifically, we describe a computational strategy, based on generative (e.g., hierarchical Bayesian) models of behavioral and functional neuroimaging data, for establishing diagnostic tests. These tests could provide estimates of specific cognitive processes underlying ASD and delineate pathophysiological mechanisms with concrete treatment targets. Written with a clinical audience in mind, this article outlines how the development of computational diagnostics applicable to behavioral and functional neuroimaging data in routine clinical practice could not only fundamentally alter our concept of ASD but eventually also transform the clinical management of this disorder.

Can Bayesian Theories of Autism Spectrum Disorder Help Improve Clinical Practice?

PubMed Central

Haker, Helene; Schneebeli, Maya; Stephan, Klaas Enno

2016-01-01

Diagnosis and individualized treatment of autism spectrum disorder (ASD) represent major problems for contemporary psychiatry. Tackling these problems requires guidance by a pathophysiological theory. In this paper, we consider recent theories that re-conceptualize ASD from a “Bayesian brain” perspective, which posit that the core abnormality of ASD resides in perceptual aberrations due to a disbalance in the precision of prediction errors (sensory noise) relative to the precision of predictions (prior beliefs). This results in percepts that are dominated by sensory inputs and less guided by top-down regularization and shifts the perceptual focus to detailed aspects of the environment with difficulties in extracting meaning. While these Bayesian theories have inspired ongoing empirical studies, their clinical implications have not yet been carved out. Here, we consider how this Bayesian perspective on disease mechanisms in ASD might contribute to improving clinical care for affected individuals. Specifically, we describe a computational strategy, based on generative (e.g., hierarchical Bayesian) models of behavioral and functional neuroimaging data, for establishing diagnostic tests. These tests could provide estimates of specific cognitive processes underlying ASD and delineate pathophysiological mechanisms with concrete treatment targets. Written with a clinical audience in mind, this article outlines how the development of computational diagnostics applicable to behavioral and functional neuroimaging data in routine clinical practice could not only fundamentally alter our concept of ASD but eventually also transform the clinical management of this disorder. PMID:27378955

Trending in Probability of Collision Measurements via a Bayesian Zero-Inflated Beta Mixed Model

NASA Technical Reports Server (NTRS)

Vallejo, Jonathon; Hejduk, Matt; Stamey, James

2015-01-01

We investigate the performance of a generalized linear mixed model in predicting the Probabilities of Collision (Pc) for conjunction events. Specifically, we apply this model to the log(sub 10) transformation of these probabilities and argue that this transformation yields values that can be considered bounded in practice. Additionally, this bounded random variable, after scaling, is zero-inflated. Consequently, we model these values using the zero-inflated Beta distribution, and utilize the Bayesian paradigm and the mixed model framework to borrow information from past and current events. This provides a natural way to model the data and provides a basis for answering questions of interest, such as what is the likelihood of observing a probability of collision equal to the effective value of zero on a subsequent observation.

Technical Note: Approximate Bayesian parameterization of a complex tropical forest model

NASA Astrophysics Data System (ADS)

Hartig, F.; Dislich, C.; Wiegand, T.; Huth, A.

2013-08-01

Inverse parameter estimation of process-based models is a long-standing problem in ecology and evolution. A key problem of inverse parameter estimation is to define a metric that quantifies how well model predictions fit to the data. Such a metric can be expressed by general cost or objective functions, but statistical inversion approaches are based on a particular metric, the probability of observing the data given the model, known as the likelihood. Deriving likelihoods for dynamic models requires making assumptions about the probability for observations to deviate from mean model predictions. For technical reasons, these assumptions are usually derived without explicit consideration of the processes in the simulation. Only in recent years have new methods become available that allow generating likelihoods directly from stochastic simulations. Previous applications of these approximate Bayesian methods have concentrated on relatively simple models. Here, we report on the application of a simulation-based likelihood approximation for FORMIND, a parameter-rich individual-based model of tropical forest dynamics. We show that approximate Bayesian inference, based on a parametric likelihood approximation placed in a conventional MCMC, performs well in retrieving known parameter values from virtual field data generated by the forest model. We analyze the results of the parameter estimation, examine the sensitivity towards the choice and aggregation of model outputs and observed data (summary statistics), and show results from using this method to fit the FORMIND model to field data from an Ecuadorian tropical forest. Finally, we discuss differences of this approach to Approximate Bayesian Computing (ABC), another commonly used method to generate simulation-based likelihood approximations. Our results demonstrate that simulation-based inference, which offers considerable conceptual advantages over more traditional methods for inverse parameter estimation, can successfully be applied to process-based models of high complexity. The methodology is particularly suited to heterogeneous and complex data structures and can easily be adjusted to other model types, including most stochastic population and individual-based models. Our study therefore provides a blueprint for a fairly general approach to parameter estimation of stochastic process-based models in ecology and evolution.

Including non-additive genetic effects in Bayesian methods for the prediction of genetic values based on genome-wide markers

PubMed Central

2011-01-01

Background Molecular marker information is a common source to draw inferences about the relationship between genetic and phenotypic variation. Genetic effects are often modelled as additively acting marker allele effects. The true mode of biological action can, of course, be different from this plain assumption. One possibility to better understand the genetic architecture of complex traits is to include intra-locus (dominance) and inter-locus (epistasis) interaction of alleles as well as the additive genetic effects when fitting a model to a trait. Several Bayesian MCMC approaches exist for the genome-wide estimation of genetic effects with high accuracy of genetic value prediction. Including pairwise interaction for thousands of loci would probably go beyond the scope of such a sampling algorithm because then millions of effects are to be estimated simultaneously leading to months of computation time. Alternative solving strategies are required when epistasis is studied. Methods We extended a fast Bayesian method (fBayesB), which was previously proposed for a purely additive model, to include non-additive effects. The fBayesB approach was used to estimate genetic effects on the basis of simulated datasets. Different scenarios were simulated to study the loss of accuracy of prediction, if epistatic effects were not simulated but modelled and vice versa. Results If 23 QTL were simulated to cause additive and dominance effects, both fBayesB and a conventional MCMC sampler BayesB yielded similar results in terms of accuracy of genetic value prediction and bias of variance component estimation based on a model including additive and dominance effects. Applying fBayesB to data with epistasis, accuracy could be improved by 5% when all pairwise interactions were modelled as well. The accuracy decreased more than 20% if genetic variation was spread over 230 QTL. In this scenario, accuracy based on modelling only additive and dominance effects was generally superior to that of the complex model including epistatic effects. Conclusions This simulation study showed that the fBayesB approach is convenient for genetic value prediction. Jointly estimating additive and non-additive effects (especially dominance) has reasonable impact on the accuracy of prediction and the proportion of genetic variation assigned to the additive genetic source. PMID:21867519

A simplified gross primary production and evapotranspiration model for boreal coniferous forests - is a generic calibration sufficient?

NASA Astrophysics Data System (ADS)

Minunno, F.; Peltoniemi, M.; Launiainen, S.; Aurela, M.; Lindroth, A.; Lohila, A.; Mammarella, I.; Minkkinen, K.; Mäkelä, A.

2015-07-01

The problem of model complexity has been lively debated in environmental sciences as well as in the forest modelling community. Simple models are less input demanding and their calibration involves a lower number of parameters, but they might be suitable only at local scale. In this work we calibrated a simplified ecosystem process model (PRELES) to data from multiple sites and we tested if PRELES can be used at regional scale to estimate the carbon and water fluxes of Boreal conifer forests. We compared a multi-site (M-S) with site-specific (S-S) calibrations. Model calibrations and evaluations were carried out by the means of the Bayesian method; Bayesian calibration (BC) and Bayesian model comparison (BMC) were used to quantify the uncertainty in model parameters and model structure. To evaluate model performances BMC results were combined with more classical analysis of model-data mismatch (M-DM). Evapotranspiration (ET) and gross primary production (GPP) measurements collected in 10 sites of Finland and Sweden were used in the study. Calibration results showed that similar estimates were obtained for the parameters at which model outputs are most sensitive. No significant differences were encountered in the predictions of the multi-site and site-specific versions of PRELES with exception of a site with agricultural history (Alkkia). Although PRELES predicted GPP better than evapotranspiration, we concluded that the model can be reliably used at regional scale to simulate carbon and water fluxes of Boreal forests. Our analyses underlined also the importance of using long and carefully collected flux datasets in model calibration. In fact, even a single site can provide model calibrations that can be applied at a wider spatial scale, since it covers a wide range of variability in climatic conditions.

Using Latent Class Analysis to Model Temperament Types

ERIC Educational Resources Information Center

Loken, Eric

2004-01-01

Mixture models are appropriate for data that arise from a set of qualitatively different subpopulations. In this study, latent class analysis was applied to observational data from a laboratory assessment of infant temperament at four months of age. The EM algorithm was used to fit the models, and the Bayesian method of posterior predictive checks…

Bayesian calibration of the Community Land Model using surrogates

DOE Office of Scientific and Technical Information (OSTI.GOV)

Ray, Jaideep; Hou, Zhangshuan; Huang, Maoyi

2014-02-01

We present results from the Bayesian calibration of hydrological parameters of the Community Land Model (CLM), which is often used in climate simulations and Earth system models. A statistical inverse problem is formulated for three hydrological parameters, conditional on observations of latent heat surface fluxes over 48 months. Our calibration method uses polynomial and Gaussian process surrogates of the CLM, and solves the parameter estimation problem using a Markov chain Monte Carlo sampler. Posterior probability densities for the parameters are developed for two sites with different soil and vegetation covers. Our method also allows us to examine the structural errormore » in CLM under two error models. We find that surrogate models can be created for CLM in most cases. The posterior distributions are more predictive than the default parameter values in CLM. Climatologically averaging the observations does not modify the parameters' distributions significantly. The structural error model reveals a correlation time-scale which can be used to identify the physical process that could be contributing to it. While the calibrated CLM has a higher predictive skill, the calibration is under-dispersive.« less

A Bayesian nonparametric method for prediction in EST analysis

PubMed Central

Lijoi, Antonio; Mena, Ramsés H; Prünster, Igor

2007-01-01

Background Expressed sequence tags (ESTs) analyses are a fundamental tool for gene identification in organisms. Given a preliminary EST sample from a certain library, several statistical prediction problems arise. In particular, it is of interest to estimate how many new genes can be detected in a future EST sample of given size and also to determine the gene discovery rate: these estimates represent the basis for deciding whether to proceed sequencing the library and, in case of a positive decision, a guideline for selecting the size of the new sample. Such information is also useful for establishing sequencing efficiency in experimental design and for measuring the degree of redundancy of an EST library. Results In this work we propose a Bayesian nonparametric approach for tackling statistical problems related to EST surveys. In particular, we provide estimates for: a) the coverage, defined as the proportion of unique genes in the library represented in the given sample of reads; b) the number of new unique genes to be observed in a future sample; c) the discovery rate of new genes as a function of the future sample size. The Bayesian nonparametric model we adopt conveys, in a statistically rigorous way, the available information into prediction. Our proposal has appealing properties over frequentist nonparametric methods, which become unstable when prediction is required for large future samples. EST libraries, previously studied with frequentist methods, are analyzed in detail. Conclusion The Bayesian nonparametric approach we undertake yields valuable tools for gene capture and prediction in EST libraries. The estimators we obtain do not feature the kind of drawbacks associated with frequentist estimators and are reliable for any size of the additional sample. PMID:17868445

A Dynamic Bayesian Network model for long-term simulation of clinical complications in type 1 diabetes.

PubMed

Marini, Simone; Trifoglio, Emanuele; Barbarini, Nicola; Sambo, Francesco; Di Camillo, Barbara; Malovini, Alberto; Manfrini, Marco; Cobelli, Claudio; Bellazzi, Riccardo

2015-10-01

The increasing prevalence of diabetes and its related complications is raising the need for effective methods to predict patient evolution and for stratifying cohorts in terms of risk of developing diabetes-related complications. In this paper, we present a novel approach to the simulation of a type 1 diabetes population, based on Dynamic Bayesian Networks, which combines literature knowledge with data mining of a rich longitudinal cohort of type 1 diabetes patients, the DCCT/EDIC study. In particular, in our approach we simulate the patient health state and complications through discretized variables. Two types of models are presented, one entirely learned from the data and the other partially driven by literature derived knowledge. The whole cohort is simulated for fifteen years, and the simulation error (i.e. for each variable, the percentage of patients predicted in the wrong state) is calculated every year on independent test data. For each variable, the population predicted in the wrong state is below 10% on both models over time. Furthermore, the distributions of real vs. simulated patients greatly overlap. Thus, the proposed models are viable tools to support decision making in type 1 diabetes. Copyright © 2015 Elsevier Inc. All rights reserved.

Optimal visuotactile integration for velocity discrimination of self-hand movements

PubMed Central

Chancel, M.; Blanchard, C.; Guerraz, M.; Montagnini, A.

2016-01-01

Illusory hand movements can be elicited by a textured disk or a visual pattern rotating under one's hand, while proprioceptive inputs convey immobility information (Blanchard C, Roll R, Roll JP, Kavounoudias A. PLoS One 8: e62475, 2013). Here, we investigated whether visuotactile integration can optimize velocity discrimination of illusory hand movements in line with Bayesian predictions. We induced illusory movements in 15 volunteers by visual and/or tactile stimulation delivered at six angular velocities. Participants had to compare hand illusion velocities with a 5°/s hand reference movement in an alternative forced choice paradigm. Results showed that the discrimination threshold decreased in the visuotactile condition compared with unimodal (visual or tactile) conditions, reflecting better bimodal discrimination. The perceptual strength (gain) of the illusions also increased: the stimulation required to give rise to a 5°/s illusory movement was slower in the visuotactile condition compared with each of the two unimodal conditions. The maximum likelihood estimation model satisfactorily predicted the improved discrimination threshold but not the increase in gain. When we added a zero-centered prior, reflecting immobility information, the Bayesian model did actually predict the gain increase but systematically overestimated it. Interestingly, the predicted gains better fit the visuotactile performances when a proprioceptive noise was generated by covibrating antagonist wrist muscles. These findings show that kinesthetic information of visual and tactile origins is optimally integrated to improve velocity discrimination of self-hand movements. However, a Bayesian model alone could not fully describe the illusory phenomenon pointing to the crucial importance of the omnipresent muscle proprioceptive cues with respect to other sensory cues for kinesthesia. PMID:27385802

Flood quantile estimation at ungauged sites by Bayesian networks

NASA Astrophysics Data System (ADS)

Mediero, L.; Santillán, D.; Garrote, L.

2012-04-01

Estimating flood quantiles at a site for which no observed measurements are available is essential for water resources planning and management. Ungauged sites have no observations about the magnitude of floods, but some site and basin characteristics are known. The most common technique used is the multiple regression analysis, which relates physical and climatic basin characteristic to flood quantiles. Regression equations are fitted from flood frequency data and basin characteristics at gauged sites. Regression equations are a rigid technique that assumes linear relationships between variables and cannot take the measurement errors into account. In addition, the prediction intervals are estimated in a very simplistic way from the variance of the residuals in the estimated model. Bayesian networks are a probabilistic computational structure taken from the field of Artificial Intelligence, which have been widely and successfully applied to many scientific fields like medicine and informatics, but application to the field of hydrology is recent. Bayesian networks infer the joint probability distribution of several related variables from observations through nodes, which represent random variables, and links, which represent causal dependencies between them. A Bayesian network is more flexible than regression equations, as they capture non-linear relationships between variables. In addition, the probabilistic nature of Bayesian networks allows taking the different sources of estimation uncertainty into account, as they give a probability distribution as result. A homogeneous region in the Tagus Basin was selected as case study. A regression equation was fitted taking the basin area, the annual maximum 24-hour rainfall for a given recurrence interval and the mean height as explanatory variables. Flood quantiles at ungauged sites were estimated by Bayesian networks. Bayesian networks need to be learnt from a huge enough data set. As observational data are reduced, a stochastic generator of synthetic data was developed. Synthetic basin characteristics were randomised, keeping the statistical properties of observed physical and climatic variables in the homogeneous region. The synthetic flood quantiles were stochastically generated taking the regression equation as basis. The learnt Bayesian network was validated by the reliability diagram, the Brier Score and the ROC diagram, which are common measures used in the validation of probabilistic forecasts. Summarising, the flood quantile estimations through Bayesian networks supply information about the prediction uncertainty as a probability distribution function of discharges is given as result. Therefore, the Bayesian network model has application as a decision support for water resources and planning management.

A Bayesian hierarchical approach to comparative audit for carotid surgery.

PubMed

Kuhan, G; Marshall, E C; Abidia, A F; Chetter, I C; McCollum, P T

2002-12-01

the aim of this study was to illustrate how a Bayesian hierarchical modelling approach can aid the reliable comparison of outcome rates between surgeons. retrospective analysis of prospective and retrospective data. binary outcome data (death/stroke within 30 days), together with information on 15 possible risk factors specific for CEA were available on 836 CEAs performed by four vascular surgeons from 1992-99. The median patient age was 68 (range 38-86) years and 60% were men. the model was developed using the WinBUGS software. After adjusting for patient-level risk factors, a cross-validatory approach was adopted to identify "divergent" performance. A ranking exercise was also carried out. the overall observed 30-day stroke/death rate was 3.9% (33/836). The model found diabetes, stroke and heart disease to be significant risk factors. There was no significant difference between the predicted and observed outcome rates for any surgeon (Bayesian p -value>0.05). Each surgeon had a median rank of 3 with associated 95% CI 1.0-5.0, despite the variability of observed stroke/death rate from 2.9-4.4%. After risk adjustment, there was very little residual between-surgeon variability in outcome rate. Bayesian hierarchical models can help to accurately quantify the uncertainty associated with surgeons' performance and rank.

Probabilistic mapping of descriptive health status responses onto health state utilities using Bayesian networks: an empirical analysis converting SF-12 into EQ-5D utility index in a national US sample.

PubMed

Le, Quang A; Doctor, Jason N

2011-05-01

As quality-adjusted life years have become the standard metric in health economic evaluations, mapping health-profile or disease-specific measures onto preference-based measures to obtain quality-adjusted life years has become a solution when health utilities are not directly available. However, current mapping methods are limited due to their predictive validity, reliability, and/or other methodological issues. We employ probability theory together with a graphical model, called a Bayesian network, to convert health-profile measures into preference-based measures and to compare the results to those estimated with current mapping methods. A sample of 19,678 adults who completed both the 12-item Short Form Health Survey (SF-12v2) and EuroQoL 5D (EQ-5D) questionnaires from the 2003 Medical Expenditure Panel Survey was split into training and validation sets. Bayesian networks were constructed to explore the probabilistic relationships between each EQ-5D domain and 12 items of the SF-12v2. The EQ-5D utility scores were estimated on the basis of the predicted probability of each response level of the 5 EQ-5D domains obtained from the Bayesian inference process using the following methods: Monte Carlo simulation, expected utility, and most-likely probability. Results were then compared with current mapping methods including multinomial logistic regression, ordinary least squares, and censored least absolute deviations. The Bayesian networks consistently outperformed other mapping models in the overall sample (mean absolute error=0.077, mean square error=0.013, and R overall=0.802), in different age groups, number of chronic conditions, and ranges of the EQ-5D index. Bayesian networks provide a new robust and natural approach to map health status responses into health utility measures for health economic evaluations.

Multi-model ensemble hydrologic prediction using Bayesian model averaging

NASA Astrophysics Data System (ADS)

Duan, Qingyun; Ajami, Newsha K.; Gao, Xiaogang; Sorooshian, Soroosh

2007-05-01

Multi-model ensemble strategy is a means to exploit the diversity of skillful predictions from different models. This paper studies the use of Bayesian model averaging (BMA) scheme to develop more skillful and reliable probabilistic hydrologic predictions from multiple competing predictions made by several hydrologic models. BMA is a statistical procedure that infers consensus predictions by weighing individual predictions based on their probabilistic likelihood measures, with the better performing predictions receiving higher weights than the worse performing ones. Furthermore, BMA provides a more reliable description of the total predictive uncertainty than the original ensemble, leading to a sharper and better calibrated probability density function (PDF) for the probabilistic predictions. In this study, a nine-member ensemble of hydrologic predictions was used to test and evaluate the BMA scheme. This ensemble was generated by calibrating three different hydrologic models using three distinct objective functions. These objective functions were chosen in a way that forces the models to capture certain aspects of the hydrograph well (e.g., peaks, mid-flows and low flows). Two sets of numerical experiments were carried out on three test basins in the US to explore the best way of using the BMA scheme. In the first set, a single set of BMA weights was computed to obtain BMA predictions, while the second set employed multiple sets of weights, with distinct sets corresponding to different flow intervals. In both sets, the streamflow values were transformed using Box-Cox transformation to ensure that the probability distribution of the prediction errors is approximately Gaussian. A split sample approach was used to obtain and validate the BMA predictions. The test results showed that BMA scheme has the advantage of generating more skillful and equally reliable probabilistic predictions than original ensemble. The performance of the expected BMA predictions in terms of daily root mean square error (DRMS) and daily absolute mean error (DABS) is generally superior to that of the best individual predictions. Furthermore, the BMA predictions employing multiple sets of weights are generally better than those using single set of weights.

Maritime Transportation Risk Assessment of Tianjin Port with Bayesian Belief Networks.

PubMed

Zhang, Jinfen; Teixeira, Ângelo P; Guedes Soares, C; Yan, Xinping; Liu, Kezhong

2016-06-01

This article develops a Bayesian belief network model for the prediction of accident consequences in the Tianjin port. The study starts with a statistical analysis of historical accident data of six years from 2008 to 2013. Then a Bayesian belief network is constructed to express the dependencies between the indicator variables and accident consequences. The statistics and expert knowledge are synthesized in the Bayesian belief network model to obtain the probability distribution of the consequences. By a sensitivity analysis, several indicator variables that have influence on the consequences are identified, including navigational area, ship type and time of the day. The results indicate that the consequences are most sensitive to the position where the accidents occurred, followed by time of day and ship length. The results also reflect that the navigational risk of the Tianjin port is at the acceptable level, despite that there is more room of improvement. These results can be used by the Maritime Safety Administration to take effective measures to enhance maritime safety in the Tianjin port. © 2016 Society for Risk Analysis.

Estimating the Uncertain Mathematical Structure of Hydrological Model via Bayesian Data Assimilation

NASA Astrophysics Data System (ADS)

Bulygina, N.; Gupta, H.; O'Donell, G.; Wheater, H.

2008-12-01

The structure of hydrological model at macro scale (e.g. watershed) is inherently uncertain due to many factors, including the lack of a robust hydrological theory at the macro scale. In this work, we assume that a suitable conceptual model for the hydrologic system has already been determined - i.e., the system boundaries have been specified, the important state variables and input and output fluxes to be included have been selected, and the major hydrological processes and geometries of their interconnections have been identified. The structural identification problem then is to specify the mathematical form of the relationships between the inputs, state variables and outputs, so that a computational model can be constructed for making simulations and/or predictions of system input-state-output behaviour. We show how Bayesian data assimilation can be used to merge both prior beliefs in the form of pre-assumed model equations with information derived from the data to construct a posterior model. The approach, entitled Bayesian Estimation of Structure (BESt), is used to estimate a hydrological model for a small basin in England, at hourly time scales, conditioned on the assumption of 3-dimensional state - soil moisture storage, fast and slow flow stores - conceptual model structure. Inputs to the system are precipitation and potential evapotranspiration, and outputs are actual evapotranspiration and streamflow discharge. Results show the difference between prior and posterior mathematical structures, as well as provide prediction confidence intervals that reflect three types of uncertainty: due to initial conditions, due to input and due to mathematical structure.

Parameter Estimation for a Turbulent Buoyant Jet Using Approximate Bayesian Computation

NASA Astrophysics Data System (ADS)

Christopher, Jason D.; Wimer, Nicholas T.; Hayden, Torrey R. S.; Lapointe, Caelan; Grooms, Ian; Rieker, Gregory B.; Hamlington, Peter E.

2016-11-01

Approximate Bayesian Computation (ABC) is a powerful tool that allows sparse experimental or other "truth" data to be used for the prediction of unknown model parameters in numerical simulations of real-world engineering systems. In this presentation, we introduce the ABC approach and then use ABC to predict unknown inflow conditions in simulations of a two-dimensional (2D) turbulent, high-temperature buoyant jet. For this test case, truth data are obtained from a simulation with known boundary conditions and problem parameters. Using spatially-sparse temperature statistics from the 2D buoyant jet truth simulation, we show that the ABC method provides accurate predictions of the true jet inflow temperature. The success of the ABC approach in the present test suggests that ABC is a useful and versatile tool for engineering fluid dynamics research.

A Bayesian network based framework for real-time crash prediction on the basic freeway segments of urban expressways.

PubMed

Hossain, Moinul; Muromachi, Yasunori

2012-03-01

The concept of measuring the crash risk for a very short time window in near future is gaining more practicality due to the recent advancements in the fields of information systems and traffic sensor technology. Although some real-time crash prediction models have already been proposed, they are still primitive in nature and require substantial improvements to be implemented in real-life. This manuscript investigates the major shortcomings of the existing models and offers solutions to overcome them with an improved framework and modeling method. It employs random multinomial logit model to identify the most important predictors as well as the most suitable detector locations to acquire data to build such a model. Afterwards, it applies Bayesian belief net (BBN) to build the real-time crash prediction model. The model has been constructed using high resolution detector data collected from Shibuya 3 and Shinjuku 4 expressways under the jurisdiction of Tokyo Metropolitan Expressway Company Limited, Japan. It has been specifically built for the basic freeway segments and it predicts the chance of formation of a hazardous traffic condition within the next 4-9 min for a particular 250 meter long road section. The performance evaluation results reflect that at an average threshold value the model is able to successful classify 66% of the future crashes with a false alarm rate less than 20%. Copyright © 2011 Elsevier Ltd. All rights reserved.

Bayesian learning of visual chunks by human observers

PubMed Central

Orbán, Gergő; Fiser, József; Aslin, Richard N.; Lengyel, Máté

2008-01-01

Efficient and versatile processing of any hierarchically structured information requires a learning mechanism that combines lower-level features into higher-level chunks. We investigated this chunking mechanism in humans with a visual pattern-learning paradigm. We developed an ideal learner based on Bayesian model comparison that extracts and stores only those chunks of information that are minimally sufficient to encode a set of visual scenes. Our ideal Bayesian chunk learner not only reproduced the results of a large set of previous empirical findings in the domain of human pattern learning but also made a key prediction that we confirmed experimentally. In accordance with Bayesian learning but contrary to associative learning, human performance was well above chance when pair-wise statistics in the exemplars contained no relevant information. Thus, humans extract chunks from complex visual patterns by generating accurate yet economical representations and not by encoding the full correlational structure of the input. PMID:18268353

Bayesian analysis of physiologically based toxicokinetic and toxicodynamic models.

PubMed

Hack, C Eric

2006-04-17

Physiologically based toxicokinetic (PBTK) and toxicodynamic (TD) models of bromate in animals and humans would improve our ability to accurately estimate the toxic doses in humans based on available animal studies. These mathematical models are often highly parameterized and must be calibrated in order for the model predictions of internal dose to adequately fit the experimentally measured doses. Highly parameterized models are difficult to calibrate and it is difficult to obtain accurate estimates of uncertainty or variability in model parameters with commonly used frequentist calibration methods, such as maximum likelihood estimation (MLE) or least squared error approaches. The Bayesian approach called Markov chain Monte Carlo (MCMC) analysis can be used to successfully calibrate these complex models. Prior knowledge about the biological system and associated model parameters is easily incorporated in this approach in the form of prior parameter distributions, and the distributions are refined or updated using experimental data to generate posterior distributions of parameter estimates. The goal of this paper is to give the non-mathematician a brief description of the Bayesian approach and Markov chain Monte Carlo analysis, how this technique is used in risk assessment, and the issues associated with this approach.

Knowledge and implicature: modeling language understanding as social cognition.

PubMed

Goodman, Noah D; Stuhlmüller, Andreas

2013-01-01

Is language understanding a special case of social cognition? To help evaluate this view, we can formalize it as the rational speech-act theory: Listeners assume that speakers choose their utterances approximately optimally, and listeners interpret an utterance by using Bayesian inference to "invert" this model of the speaker. We apply this framework to model scalar implicature ("some" implies "not all," and "N" implies "not more than N"). This model predicts an interaction between the speaker's knowledge state and the listener's interpretation. We test these predictions in two experiments and find good fit between model predictions and human judgments. Copyright © 2013 Cognitive Science Society, Inc.

IMAGINE: Interstellar MAGnetic field INference Engine

NASA Astrophysics Data System (ADS)

Steininger, Theo

2018-03-01

IMAGINE (Interstellar MAGnetic field INference Engine) performs inference on generic parametric models of the Galaxy. The modular open source framework uses highly optimized tools and technology such as the MultiNest sampler (ascl:1109.006) and the information field theory framework NIFTy (ascl:1302.013) to create an instance of the Milky Way based on a set of parameters for physical observables, using Bayesian statistics to judge the mismatch between measured data and model prediction. The flexibility of the IMAGINE framework allows for simple refitting for newly available data sets and makes state-of-the-art Bayesian methods easily accessible particularly for random components of the Galactic magnetic field.

Generation of Bayesian prediction models for OATP-mediated drug-drug interactions based on inhibition screen of OATP1B1, OATP1B1∗15 and OATP1B3.

PubMed

van de Steeg, E; Venhorst, J; Jansen, H T; Nooijen, I H G; DeGroot, J; Wortelboer, H M; Vlaming, M L H

2015-04-05

Human organic anion-transporting polypeptide 1B1 (OATP1B1) and OATP1B3 are important hepatic uptake transporters. Early assessment of OATP1B1/1B3-mediated drug-drug interactions (DDIs) is therefore important for successful drug development. A promising approach for early screening and prediction of DDIs is computational modeling. In this study we aimed to generate a rapid, single Bayesian prediction model for OATP1B1, OATP1B1∗15 and OATP1B3 inhibition. Besides our previously generated HEK-OATP1B1 and HEK-OATP1B1∗15 cells, we now generated and characterized HEK-OATP1B3 cells. Using these cell lines we investigated the inhibitory potential of 640 FDA-approved drugs from a commercial library (10μM) on the uptake of [(3)H]-estradiol-17β-d-glucuronide (1μM) by OATP1B1, OATP1B1∗15, and OATP1B3. Using a cut-off of ⩾60% inhibition, 8% and 7% of the 640 drugs were potent OATP1B1 and OATP1B1∗15 inhibitors, respectively. Only 1% of the tested drugs significantly inhibited OATP1B3, which was not sufficient for Bayesian modeling. Modeling of OATP1B1 and OATP1B1∗15 inhibition revealed that presence of conjugated systems and (hetero)cycles with acceptor/donor atoms in- or outside the ring enhance the probability of a molecule binding these transporters. The overall performance of the model for OATP1B1 and OATP1B1∗15 was ⩾80%, including evaluation with a true external test set. Our Bayesian classification model thus represents a fast, inexpensive and robust means of assessing potential binding of new chemical entities to OATP1B1 and OATP1B1∗15. As such, this model may be used to rank compounds early in the drug development process, helping to avoid adverse effects in a later stage due to inhibition of OATP1B1 and/or OATP1B1∗15. Copyright © 2015 Elsevier B.V. All rights reserved.

On how to avoid input and structural uncertainties corrupt the inference of hydrological parameters using a Bayesian framework

NASA Astrophysics Data System (ADS)

Hernández, Mario R.; Francés, Félix

2015-04-01

One phase of the hydrological models implementation process, significantly contributing to the hydrological predictions uncertainty, is the calibration phase in which values of the unknown model parameters are tuned by optimizing an objective function. An unsuitable error model (e.g. Standard Least Squares or SLS) introduces noise into the estimation of the parameters. The main sources of this noise are the input errors and the hydrological model structural deficiencies. Thus, the biased calibrated parameters cause the divergence model phenomenon, where the errors variance of the (spatially and temporally) forecasted flows far exceeds the errors variance in the fitting period, and provoke the loss of part or all of the physical meaning of the modeled processes. In other words, yielding a calibrated hydrological model which works well, but not for the right reasons. Besides, an unsuitable error model yields a non-reliable predictive uncertainty assessment. Hence, with the aim of prevent all these undesirable effects, this research focuses on the Bayesian joint inference (BJI) of both the hydrological and error model parameters, considering a general additive (GA) error model that allows for correlation, non-stationarity (in variance and bias) and non-normality of model residuals. As hydrological model, it has been used a conceptual distributed model called TETIS, with a particular split structure of the effective model parameters. Bayesian inference has been performed with the aid of a Markov Chain Monte Carlo (MCMC) algorithm called Dream-ZS. MCMC algorithm quantifies the uncertainty of the hydrological and error model parameters by getting the joint posterior probability distribution, conditioned on the observed flows. The BJI methodology is a very powerful and reliable tool, but it must be used correctly this is, if non-stationarity in errors variance and bias is modeled, the Total Laws must be taken into account. The results of this research show that the application of BJI with a GA error model outperforms the hydrological parameters robustness (diminishing the divergence model phenomenon) and improves the reliability of the streamflow predictive distribution, in respect of the results of a bad error model as SLS. Finally, the most likely prediction in a validation period, for both BJI+GA and SLS error models shows a similar performance.

Wind power application research on the fusion of the determination and ensemble prediction

NASA Astrophysics Data System (ADS)

Lan, Shi; Lina, Xu; Yuzhu, Hao

2017-07-01

The fused product of wind speed for the wind farm is designed through the use of wind speed products of ensemble prediction from the European Centre for Medium-Range Weather Forecasts (ECMWF) and professional numerical model products on wind power based on Mesoscale Model5 (MM5) and Beijing Rapid Update Cycle (BJ-RUC), which are suitable for short-term wind power forecasting and electric dispatch. The single-valued forecast is formed by calculating the different ensemble statistics of the Bayesian probabilistic forecasting representing the uncertainty of ECMWF ensemble prediction. Using autoregressive integrated moving average (ARIMA) model to improve the time resolution of the single-valued forecast, and based on the Bayesian model averaging (BMA) and the deterministic numerical model prediction, the optimal wind speed forecasting curve and the confidence interval are provided. The result shows that the fusion forecast has made obvious improvement to the accuracy relative to the existing numerical forecasting products. Compared with the 0-24 h existing deterministic forecast in the validation period, the mean absolute error (MAE) is decreased by 24.3 % and the correlation coefficient (R) is increased by 12.5 %. In comparison with the ECMWF ensemble forecast, the MAE is reduced by 11.7 %, and R is increased 14.5 %. Additionally, MAE did not increase with the prolongation of the forecast ahead.

Bayesian Treatment of Uncertainty in Environmental Modeling: Optimization, Sampling and Data Assimilation Using the DREAM Software Package

NASA Astrophysics Data System (ADS)

Vrugt, J. A.

2012-12-01

In the past decade much progress has been made in the treatment of uncertainty in earth systems modeling. Whereas initial approaches has focused mostly on quantification of parameter and predictive uncertainty, recent methods attempt to disentangle the effects of parameter, forcing (input) data, model structural and calibration data errors. In this talk I will highlight some of our recent work involving theory, concepts and applications of Bayesian parameter and/or state estimation. In particular, new methods for sequential Monte Carlo (SMC) and Markov Chain Monte Carlo (MCMC) simulation will be presented with emphasis on massively parallel distributed computing and quantification of model structural errors. The theoretical and numerical developments will be illustrated using model-data synthesis problems in hydrology, hydrogeology and geophysics.

Metadata Creation Tool Content Template For Data Stewards

EPA Science Inventory

A space-time Bayesian fusion model (McMillan, Holland, Morara, and Feng, 2009) is used to provide daily, gridded predictive PM_2.5 (daily average) and O₃ (daily 8-hr maximum) surfaces for 2001-2005. The fusion model uses both air quality monitoring data from ...

Why environmental scientists are becoming Bayesians

Treesearch

James S. Clark

2005-01-01

Advances in computational statistics provide a general framework for the high dimensional models typically needed for ecological inference and prediction. Hierarchical Bayes (HB) represents a modelling structure with capacity to exploit diverse sources of information, to accommodate influences that are unknown (or unknowable), and to draw inference on large numbers of...

Modelling multi-hazard hurricane damages on an urbanized coast with a Bayesian Network approach

USGS Publications Warehouse

van Verseveld, H.C.W.; Van Dongeren, A. R.; Plant, Nathaniel G.; Jäger, W.S.; den Heijer, C.

2015-01-01

Hurricane flood impacts to residential buildings in coastal zones are caused by a number of hazards, such as inundation, overflow currents, erosion, and wave attack. However, traditional hurricane damage models typically make use of stage-damage functions, where the stage is related to flooding depth only. Moreover, these models are deterministic and do not consider the large amount of uncertainty associated with both the processes themselves and with the predictions. This uncertainty becomes increasingly important when multiple hazards (flooding, wave attack, erosion, etc.) are considered simultaneously. This paper focusses on establishing relationships between observed damage and multiple hazard indicators in order to make better probabilistic predictions. The concept consists of (1) determining Local Hazard Indicators (LHIs) from a hindcasted storm with use of a nearshore morphodynamic model, XBeach, and (2) coupling these LHIs and building characteristics to the observed damages. We chose a Bayesian Network approach in order to make this coupling and used the LHIs ‘Inundation depth’, ‘Flow velocity’, ‘Wave attack’, and ‘Scour depth’ to represent flooding, current, wave impacts, and erosion related hazards.The coupled hazard model was tested against four thousand damage observations from a case site at the Rockaway Peninsula, NY, that was impacted by Hurricane Sandy in late October, 2012. The model was able to accurately distinguish ‘Minor damage’ from all other outcomes 95% of the time and could distinguish areas that were affected by the storm, but not severely damaged, 68% of the time. For the most heavily damaged buildings (‘Major Damage’ and ‘Destroyed’), projections of the expected damage underestimated the observed damage. The model demonstrated that including multiple hazards doubled the prediction skill, with Log-Likelihood Ratio test (a measure of improved accuracy and reduction in uncertainty) scores between 0.02 and 0.17 when only one hazard is considered and a score of 0.37 when multiple hazards are considered simultaneously. The LHIs with the most predictive skill were ‘Inundation depth’ and ‘Wave attack’. The Bayesian Network approach has several advantages over the market-standard stage-damage functions: the predictive capacity of multiple indicators can be combined; probabilistic predictions can be obtained, which include uncertainty; and quantitative as well as descriptive information can be used simultaneously.

Uncertainty quantification for nuclear density functional theory and information content of new measurements

DOE PAGES

McDonnell, J. D.; Schunck, N.; Higdon, D.; ...

2015-03-24

Statistical tools of uncertainty quantification can be used to assess the information content of measured observables with respect to present-day theoretical models, to estimate model errors and thereby improve predictive capability, to extrapolate beyond the regions reached by experiment, and to provide meaningful input to applications and planned measurements. To showcase new opportunities offered by such tools, we make a rigorous analysis of theoretical statistical uncertainties in nuclear density functional theory using Bayesian inference methods. By considering the recent mass measurements from the Canadian Penning Trap at Argonne National Laboratory, we demonstrate how the Bayesian analysis and a direct least-squaresmore » optimization, combined with high-performance computing, can be used to assess the information content of the new data with respect to a model based on the Skyrme energy density functional approach. Employing the posterior probability distribution computed with a Gaussian process emulator, we apply the Bayesian framework to propagate theoretical statistical uncertainties in predictions of nuclear masses, two-neutron dripline, and fission barriers. Overall, we find that the new mass measurements do not impose a constraint that is strong enough to lead to significant changes in the model parameters. In addition, the example discussed in this study sets the stage for quantifying and maximizing the impact of new measurements with respect to current modeling and guiding future experimental efforts, thus enhancing the experiment-theory cycle in the scientific method.« less

Uncertainty quantification for nuclear density functional theory and information content of new measurements

DOE Office of Scientific and Technical Information (OSTI.GOV)

McDonnell, J. D.; Schunck, N.; Higdon, D.

2015-03-24

Statistical tools of uncertainty quantification can be used to assess the information content of measured observables with respect to present-day theoretical models, to estimate model errors and thereby improve predictive capability, to extrapolate beyond the regions reached by experiment, and to provide meaningful input to applications and planned measurements. To showcase new opportunities offered by such tools, we make a rigorous analysis of theoretical statistical uncertainties in nuclear density functional theory using Bayesian inference methods. By considering the recent mass measurements from the Canadian Penning Trap at Argonne National Laboratory, we demonstrate how the Bayesian analysis and a direct least-squaresmore » optimization, combined with high-performance computing, can be used to assess the information content of the new data with respect to a model based on the Skyrme energy density functional approach. Employing the posterior probability distribution computed with a Gaussian process emulator, we apply the Bayesian framework to propagate theoretical statistical uncertainties in predictions of nuclear masses, two-neutron dripline, and fission barriers. Overall, we find that the new mass measurements do not impose a constraint that is strong enough to lead to significant changes in the model parameters. As a result, the example discussed in this study sets the stage for quantifying and maximizing the impact of new measurements with respect to current modeling and guiding future experimental efforts, thus enhancing the experiment-theory cycle in the scientific method.« less

Spatial Prediction and Optimized Sampling Design for Sodium Concentration in Groundwater

PubMed Central

Shabbir, Javid; M. AbdEl-Salam, Nasser; Hussain, Tajammal

2016-01-01

Sodium is an integral part of water, and its excessive amount in drinking water causes high blood pressure and hypertension. In the present paper, spatial distribution of sodium concentration in drinking water is modeled and optimized sampling designs for selecting sampling locations is calculated for three divisions in Punjab, Pakistan. Universal kriging and Bayesian universal kriging are used to predict the sodium concentrations. Spatial simulated annealing is used to generate optimized sampling designs. Different estimation methods (i.e., maximum likelihood, restricted maximum likelihood, ordinary least squares, and weighted least squares) are used to estimate the parameters of the variogram model (i.e, exponential, Gaussian, spherical and cubic). It is concluded that Bayesian universal kriging fits better than universal kriging. It is also observed that the universal kriging predictor provides minimum mean universal kriging variance for both adding and deleting locations during sampling design. PMID:27683016

Analog ensemble and Bayesian regression techniques to improve the wind speed prediction during extreme storms in the NE U.S.

NASA Astrophysics Data System (ADS)

Yang, J.; Astitha, M.; Delle Monache, L.; Alessandrini, S.

2016-12-01

Accuracy of weather forecasts in Northeast U.S. has become very important in recent years, given the serious and devastating effects of extreme weather events. Despite the use of evolved forecasting tools and techniques strengthened by increased super-computing resources, the weather forecasting systems still have their limitations in predicting extreme events. In this study, we examine the combination of analog ensemble and Bayesian regression techniques to improve the prediction of storms that have impacted NE U.S., mostly defined by the occurrence of high wind speeds (i.e. blizzards, winter storms, hurricanes and thunderstorms). The predicted wind speed, wind direction and temperature by two state-of-the-science atmospheric models (WRF and RAMS/ICLAMS) are combined using the mentioned techniques, exploring various ways that those variables influence the minimization of the prediction error (systematic and random). This study is focused on retrospective simulations of 146 storms that affected the NE U.S. in the period 2005-2016. In order to evaluate the techniques, leave-one-out cross validation procedure was implemented regarding 145 storms as the training dataset. The analog ensemble method selects a set of past observations that corresponded to the best analogs of the numerical weather prediction and provides a set of ensemble members of the selected observation dataset. The set of ensemble members can then be used in a deterministic or probabilistic way. In the Bayesian regression framework, optimal variances are estimated for the training partition by minimizing the root mean square error and are applied to the out-of-sample storm. The preliminary results indicate a significant improvement in the statistical metrics of 10-m wind speed for 146 storms using both techniques (20-30% bias and error reduction in all observation-model pairs). In this presentation, we discuss the various combinations of atmospheric predictors and techniques and illustrate how the long record of predicted storms is valuable in the improvement of wind speed prediction.

Comparison of Models and Whole-Genome Profiling Approaches for Genomic-Enabled Prediction of Septoria Tritici Blotch, Stagonospora Nodorum Blotch, and Tan Spot Resistance in Wheat.

PubMed

Juliana, Philomin; Singh, Ravi P; Singh, Pawan K; Crossa, Jose; Rutkoski, Jessica E; Poland, Jesse A; Bergstrom, Gary C; Sorrells, Mark E

2017-07-01

The leaf spotting diseases in wheat that include Septoria tritici blotch (STB) caused by , Stagonospora nodorum blotch (SNB) caused by , and tan spot (TS) caused by pose challenges to breeding programs in selecting for resistance. A promising approach that could enable selection prior to phenotyping is genomic selection that uses genome-wide markers to estimate breeding values (BVs) for quantitative traits. To evaluate this approach for seedling and/or adult plant resistance (APR) to STB, SNB, and TS, we compared the predictive ability of least-squares (LS) approach with genomic-enabled prediction models including genomic best linear unbiased predictor (GBLUP), Bayesian ridge regression (BRR), Bayes A (BA), Bayes B (BB), Bayes Cπ (BC), Bayesian least absolute shrinkage and selection operator (BL), and reproducing kernel Hilbert spaces markers (RKHS-M), a pedigree-based model (RKHS-P) and RKHS markers and pedigree (RKHS-MP). We observed that LS gave the lowest prediction accuracies and RKHS-MP, the highest. The genomic-enabled prediction models and RKHS-P gave similar accuracies. The increase in accuracy using genomic prediction models over LS was 48%. The mean genomic prediction accuracies were 0.45 for STB (APR), 0.55 for SNB (seedling), 0.66 for TS (seedling) and 0.48 for TS (APR). We also compared markers from two whole-genome profiling approaches: genotyping by sequencing (GBS) and diversity arrays technology sequencing (DArTseq) for prediction. While, GBS markers performed slightly better than DArTseq, combining markers from the two approaches did not improve accuracies. We conclude that implementing GS in breeding for these diseases would help to achieve higher accuracies and rapid gains from selection. Copyright © 2017 Crop Science Society of America.

A Bayesian network for modelling blood glucose concentration and exercise in type 1 diabetes.

PubMed

Ewings, Sean M; Sahu, Sujit K; Valletta, John J; Byrne, Christopher D; Chipperfield, Andrew J

2015-06-01

This article presents a new statistical approach to analysing the effects of everyday physical activity on blood glucose concentration in people with type 1 diabetes. A physiologically based model of blood glucose dynamics is developed to cope with frequently sampled data on food, insulin and habitual physical activity; the model is then converted to a Bayesian network to account for measurement error and variability in the physiological processes. A simulation study is conducted to determine the feasibility of using Markov chain Monte Carlo methods for simultaneous estimation of all model parameters and prediction of blood glucose concentration. Although there are problems with parameter identification in a minority of cases, most parameters can be estimated without bias. Predictive performance is unaffected by parameter misspecification and is insensitive to misleading prior distributions. This article highlights important practical and theoretical issues not previously addressed in the quest for an artificial pancreas as treatment for type 1 diabetes. The proposed methods represent a new paradigm for analysis of deterministic mathematical models of blood glucose concentration. © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav.

Understanding seasonal variability of uncertainty in hydrological prediction

NASA Astrophysics Data System (ADS)

Li, M.; Wang, Q. J.

2012-04-01

Understanding uncertainty in hydrological prediction can be highly valuable for improving the reliability of streamflow prediction. In this study, a monthly water balance model, WAPABA, in a Bayesian joint probability with error models are presented to investigate the seasonal dependency of prediction error structure. A seasonal invariant error model, analogous to traditional time series analysis, uses constant parameters for model error and account for no seasonal variations. In contrast, a seasonal variant error model uses a different set of parameters for bias, variance and autocorrelation for each individual calendar month. Potential connection amongst model parameters from similar months is not considered within the seasonal variant model and could result in over-fitting and over-parameterization. A hierarchical error model further applies some distributional restrictions on model parameters within a Bayesian hierarchical framework. An iterative algorithm is implemented to expedite the maximum a posterior (MAP) estimation of a hierarchical error model. Three error models are applied to forecasting streamflow at a catchment in southeast Australia in a cross-validation analysis. This study also presents a number of statistical measures and graphical tools to compare the predictive skills of different error models. From probability integral transform histograms and other diagnostic graphs, the hierarchical error model conforms better to reliability when compared to the seasonal invariant error model. The hierarchical error model also generally provides the most accurate mean prediction in terms of the Nash-Sutcliffe model efficiency coefficient and the best probabilistic prediction in terms of the continuous ranked probability score (CRPS). The model parameters of the seasonal variant error model are very sensitive to each cross validation, while the hierarchical error model produces much more robust and reliable model parameters. Furthermore, the result of the hierarchical error model shows that most of model parameters are not seasonal variant except for error bias. The seasonal variant error model is likely to use more parameters than necessary to maximize the posterior likelihood. The model flexibility and robustness indicates that the hierarchical error model has great potential for future streamflow predictions.

Bayesian Integration of Isotope Ratio for Geographic Sourcing of Castor Beans

DOE PAGES

Webb-Robertson, Bobbie-Jo; Kreuzer, Helen; Hart, Garret; ...

2012-01-01

Recenmore » t years have seen an increase in the forensic interest associated with the poison ricin, which is extracted from the seeds of the Ricinus communis plant. Both light element (C, N, O, and H) and strontium (Sr) isotope ratios have previously been used to associate organic material with geographic regions of origin. We present a Bayesian integration methodology that can more accurately predict the region of origin for a castor bean than individual models developed independently for light element stable isotopes or Sr isotope ratios. Our results demonstrate a clear improvement in the ability to correctly classify regions based on the integrated model with a class accuracy of 60.9 ± 2.1 % versus 55.9 ± 2.1 % and 40.2 ± 1.8 % for the light element and strontium (Sr) isotope ratios, respectively. In addition, we show graphically the strengths and weaknesses of each dataset in respect to class prediction and how the integration of these datasets strengthens the overall model.« less

Translating Uncertain Sea Level Projections Into Infrastructure Impacts Using a Bayesian Framework

NASA Astrophysics Data System (ADS)

Moftakhari, Hamed; AghaKouchak, Amir; Sanders, Brett F.; Matthew, Richard A.; Mazdiyasni, Omid

2017-12-01

Climate change may affect ocean-driven coastal flooding regimes by both raising the mean sea level (msl) and altering ocean-atmosphere interactions. For reliable projections of coastal flood risk, information provided by different climate models must be considered in addition to associated uncertainties. In this paper, we propose a framework to project future coastal water levels and quantify the resulting flooding hazard to infrastructure. We use Bayesian Model Averaging to generate a weighted ensemble of storm surge predictions from eight climate models for two coastal counties in California. The resulting ensembles combined with msl projections, and predicted astronomical tides are then used to quantify changes in the likelihood of road flooding under representative concentration pathways 4.5 and 8.5 in the near-future (1998-2063) and mid-future (2018-2083). The results show that road flooding rates will be significantly higher in the near-future and mid-future compared to the recent past (1950-2015) if adaptation measures are not implemented.

Bayesian Integration of Isotope Ratios for Geographic Sourcing of Castor Beans

DOE Office of Scientific and Technical Information (OSTI.GOV)

Webb-Robertson, Bobbie-Jo M.; Kreuzer, Helen W.; Hart, Garret L.

Recent years have seen an increase in the forensic interest associated with the poison ricin, which is extracted from the seeds of the Ricinus communis plant. Both light element (C, N, O, and H) and strontium (Sr) isotope ratios have previously been used to associate organic material with geographic regions of origin. We present a Bayesian integration methodology that can more accurately predict the region of origin for a castor bean than individual models developed independently for light element stable isotopes or Sr isotope ratios. Our results demonstrate a clear improvement in the ability to correctly classify regions based onmore » the integrated model with a class accuracy of 6 0 . 9 {+-} 2 . 1 % versus 5 5 . 9 {+-} 2 . 1 % and 4 0 . 2 {+-} 1 . 8 % for the light element and strontium (Sr) isotope ratios, respectively. In addition, we show graphically the strengths and weaknesses of each dataset in respect to class prediction and how the integration of these datasets strengthens the overall model.« less

Bayesian Integration of Isotope Ratio for Geographic Sourcing of Castor Beans

DOE Office of Scientific and Technical Information (OSTI.GOV)

Webb-Robertson, Bobbie-Jo; Kreuzer, Helen; Hart, Garret

Recenmore » t years have seen an increase in the forensic interest associated with the poison ricin, which is extracted from the seeds of the Ricinus communis plant. Both light element (C, N, O, and H) and strontium (Sr) isotope ratios have previously been used to associate organic material with geographic regions of origin. We present a Bayesian integration methodology that can more accurately predict the region of origin for a castor bean than individual models developed independently for light element stable isotopes or Sr isotope ratios. Our results demonstrate a clear improvement in the ability to correctly classify regions based on the integrated model with a class accuracy of 60.9 ± 2.1 % versus 55.9 ± 2.1 % and 40.2 ± 1.8 % for the light element and strontium (Sr) isotope ratios, respectively. In addition, we show graphically the strengths and weaknesses of each dataset in respect to class prediction and how the integration of these datasets strengthens the overall model.« less

Bayesian Integration of Isotope Ratio for Geographic Sourcing of Castor Beans

PubMed Central

Webb-Robertson, Bobbie-Jo; Kreuzer, Helen; Hart, Garret; Ehleringer, James; West, Jason; Gill, Gary; Duckworth, Douglas

2012-01-01

Recent years have seen an increase in the forensic interest associated with the poison ricin, which is extracted from the seeds of the Ricinus communis plant. Both light element (C, N, O, and H) and strontium (Sr) isotope ratios have previously been used to associate organic material with geographic regions of origin. We present a Bayesian integration methodology that can more accurately predict the region of origin for a castor bean than individual models developed independently for light element stable isotopes or Sr isotope ratios. Our results demonstrate a clear improvement in the ability to correctly classify regions based on the integrated model with a class accuracy of 60.9 ± 2.1% versus 55.9 ± 2.1% and 40.2 ± 1.8% for the light element and strontium (Sr) isotope ratios, respectively. In addition, we show graphically the strengths and weaknesses of each dataset in respect to class prediction and how the integration of these datasets strengthens the overall model. PMID:22919270

Modeling Non-Gaussian Time Series with Nonparametric Bayesian Model.

PubMed

Xu, Zhiguang; MacEachern, Steven; Xu, Xinyi

2015-02-01

We present a class of Bayesian copula models whose major components are the marginal (limiting) distribution of a stationary time series and the internal dynamics of the series. We argue that these are the two features with which an analyst is typically most familiar, and hence that these are natural components with which to work. For the marginal distribution, we use a nonparametric Bayesian prior distribution along with a cdf-inverse cdf transformation to obtain large support. For the internal dynamics, we rely on the traditionally successful techniques of normal-theory time series. Coupling the two components gives us a family of (Gaussian) copula transformed autoregressive models. The models provide coherent adjustments of time scales and are compatible with many extensions, including changes in volatility of the series. We describe basic properties of the models, show their ability to recover non-Gaussian marginal distributions, and use a GARCH modification of the basic model to analyze stock index return series. The models are found to provide better fit and improved short-range and long-range predictions than Gaussian competitors. The models are extensible to a large variety of fields, including continuous time models, spatial models, models for multiple series, models driven by external covariate streams, and non-stationary models.

Comparing vector-based and Bayesian memory models using large-scale datasets: User-generated hashtag and tag prediction on Twitter and Stack Overflow.

PubMed

Stanley, Clayton; Byrne, Michael D

2016-12-01

The growth of social media and user-created content on online sites provides unique opportunities to study models of human declarative memory. By framing the task of choosing a hashtag for a tweet and tagging a post on Stack Overflow as a declarative memory retrieval problem, 2 cognitively plausible declarative memory models were applied to millions of posts and tweets and evaluated on how accurately they predict a user's chosen tags. An ACT-R based Bayesian model and a random permutation vector-based model were tested on the large data sets. The results show that past user behavior of tag use is a strong predictor of future behavior. Furthermore, past behavior was successfully incorporated into the random permutation model that previously used only context. Also, ACT-R's attentional weight term was linked to an entropy-weighting natural language processing method used to attenuate high-frequency words (e.g., articles and prepositions). Word order was not found to be a strong predictor of tag use, and the random permutation model performed comparably to the Bayesian model without including word order. This shows that the strength of the random permutation model is not in the ability to represent word order, but rather in the way in which context information is successfully compressed. The results of the large-scale exploration show how the architecture of the 2 memory models can be modified to significantly improve accuracy, and may suggest task-independent general modifications that can help improve model fit to human data in a much wider range of domains. (PsycINFO Database Record (c) 2016 APA, all rights reserved).

Action selection performance of a reconfigurable basal ganglia inspired model with Hebbian–Bayesian Go-NoGo connectivity

PubMed Central

Berthet, Pierre; Hellgren-Kotaleski, Jeanette; Lansner, Anders

2012-01-01

Several studies have shown a strong involvement of the basal ganglia (BG) in action selection and dopamine dependent learning. The dopaminergic signal to striatum, the input stage of the BG, has been commonly described as coding a reward prediction error (RPE), i.e., the difference between the predicted and actual reward. The RPE has been hypothesized to be critical in the modulation of the synaptic plasticity in cortico-striatal synapses in the direct and indirect pathway. We developed an abstract computational model of the BG, with a dual pathway structure functionally corresponding to the direct and indirect pathways, and compared its behavior to biological data as well as other reinforcement learning models. The computations in our model are inspired by Bayesian inference, and the synaptic plasticity changes depend on a three factor Hebbian–Bayesian learning rule based on co-activation of pre- and post-synaptic units and on the value of the RPE. The model builds on a modified Actor-Critic architecture and implements the direct (Go) and the indirect (NoGo) pathway, as well as the reward prediction (RP) system, acting in a complementary fashion. We investigated the performance of the model system when different configurations of the Go, NoGo, and RP system were utilized, e.g., using only the Go, NoGo, or RP system, or combinations of those. Learning performance was investigated in several types of learning paradigms, such as learning-relearning, successive learning, stochastic learning, reversal learning and a two-choice task. The RPE and the activity of the model during learning were similar to monkey electrophysiological and behavioral data. Our results, however, show that there is not a unique best way to configure this BG model to handle well all the learning paradigms tested. We thus suggest that an agent might dynamically configure its action selection mode, possibly depending on task characteristics and also on how much time is available. PMID:23060764

Experimental design and Bayesian networks for enhancement of delta-endotoxin production by Bacillus thuringiensis.

PubMed

Ennouri, Karim; Ayed, Rayda Ben; Hassen, Hanen Ben; Mazzarello, Maura; Ottaviani, Ennio

2015-12-01

Bacillus thuringiensis (Bt) is a Gram-positive bacterium. The entomopathogenic activity of Bt is related to the existence of the crystal consisting of protoxins, also called delta-endotoxins. In order to optimize and explain the production of delta-endotoxins of Bacillus thuringiensis kurstaki, we studied seven medium components: soybean meal, starch, KH₂PO₄, K₂HPO₄, FeSO₄, MnSO₄, and MgSO₄and their relationships with the concentration of delta-endotoxins using an experimental design (Plackett-Burman design) and Bayesian networks modelling. The effects of the ingredients of the culture medium on delta-endotoxins production were estimated. The developed model showed that different medium components are important for the Bacillus thuringiensis fermentation. The most important factors influenced the production of delta-endotoxins are FeSO₄, K2HPO₄, starch and soybean meal. Indeed, it was found that soybean meal, K₂HPO₄, KH₂PO₄and starch also showed positive effect on the delta-endotoxins production. However, FeSO4 and MnSO4 expressed opposite effect. The developed model, based on Bayesian techniques, can automatically learn emerging models in data to serve in the prediction of delta-endotoxins concentrations. The constructed model in the present study implies that experimental design (Plackett-Burman design) joined with Bayesian networks method could be used for identification of effect variables on delta-endotoxins variation.

Development of a population pharmacokinetic model for carbamazepine based on sparse therapeutic monitoring data from pediatric patients with epilepsy.

PubMed

Carlsson, Kristin Cecilie; Hoem, Nils Ove; Glauser, Tracy; Vinks, Alexander A

2005-05-01

Population models can be important extensions of therapeutic drug monitoring (TDM), as they allow estimation of individual pharmacokinetic parameters based on a small number of measured drug concentrations. This study used a Bayesian approach to explore the utility of routinely collected and sparse TDM data (1 sample per patient) for carbamazepine (CBZ) monotherapy in developing a population pharmacokinetic (PPK) model for CBZ in pediatric patients that would allow prediction of CBZ concentrations for both immediate- and controlled-release formulations. Patient and TDM data were obtained from a pediatric neurology outpatient database. Data were analyzed using an iterative 2-stage Bayesian algorithm and a nonparametric adaptive grid algorithm. Models were compared by final log likelihood, mean error (ME) as a measure of bias, and root mean squared error (RMSE) as a measure of precision. Fifty-seven entries with data on CBZ monotherapy were identified from the database and used in the analysis (36 from males, 21 from females; mean [SD] age, 9.1 [4.4] years [range, 2-21 years]). Preliminary models estimating clearance (Cl) or the elimination rate constant (K(el)) gave good prediction of serum concentrations compared with measured serum concentrations, but estimates of Cl and K(el) were highly correlated with estimates of volume of distribution (V(d)). Different covariate models were then tested. The selected model had zero-order input and had age and body weight as covariates. Cl (L/h) was calculated as K(el) . V(d), where K(el) = [K(i) - (K(s) . age)] and V(d) = [V(i) + (V(s) . body weight)]. Median parameter estimates were V(i) (intercept) = 11.5 L (fixed); V(s) (slope) = 0.3957 L/kg (range, 0.01200-1.5730); K(i) (intercept) = 0.173 h(-1) (fixed); and K(s) (slope) = 0.004487 h(-1) . y(-1) (range, 0.0001800-0.02969). The fit was good for estimates of steady-state serum concentrations based on prior values (population median estimates) (R = 0.468; R(2) = 0.219) but was even better for predictions based on individual Bayesian posterior values (R(2) = 0.991), with little bias (ME = -0.079) and good precision (RMSE = 0.055). Based on the findings of this study, sparse TDM data can be used for PPK modeling of CBZ clearance in children with epilepsy, and these models can be used to predict Cl at steady state in pediatric patients. However, to estimate additional pharmacokinetic model parameters (eg, the absorption rate constant and V(d)), it would be necessary to combine sparse TDM data with additional well-timed samples. This would allow development of more informative PPK models that could be used as part of Bayesian dose-individualization strategies.

Inferring on the Intentions of Others by Hierarchical Bayesian Learning

PubMed Central

Diaconescu, Andreea O.; Mathys, Christoph; Weber, Lilian A. E.; Daunizeau, Jean; Kasper, Lars; Lomakina, Ekaterina I.; Fehr, Ernst; Stephan, Klaas E.

2014-01-01

Inferring on others' (potentially time-varying) intentions is a fundamental problem during many social transactions. To investigate the underlying mechanisms, we applied computational modeling to behavioral data from an economic game in which 16 pairs of volunteers (randomly assigned to “player” or “adviser” roles) interacted. The player performed a probabilistic reinforcement learning task, receiving information about a binary lottery from a visual pie chart. The adviser, who received more predictive information, issued an additional recommendation. Critically, the game was structured such that the adviser's incentives to provide helpful or misleading information varied in time. Using a meta-Bayesian modeling framework, we found that the players' behavior was best explained by the deployment of hierarchical learning: they inferred upon the volatility of the advisers' intentions in order to optimize their predictions about the validity of their advice. Beyond learning, volatility estimates also affected the trial-by-trial variability of decisions: participants were more likely to rely on their estimates of advice accuracy for making choices when they believed that the adviser's intentions were presently stable. Finally, our model of the players' inference predicted the players' interpersonal reactivity index (IRI) scores, explicit ratings of the advisers' helpfulness and the advisers' self-reports on their chosen strategy. Overall, our results suggest that humans (i) employ hierarchical generative models to infer on the changing intentions of others, (ii) use volatility estimates to inform decision-making in social interactions, and (iii) integrate estimates of advice accuracy with non-social sources of information. The Bayesian framework presented here can quantify individual differences in these mechanisms from simple behavioral readouts and may prove useful in future clinical studies of maladaptive social cognition. PMID:25187943

Transmembrane Topology and Signal Peptide Prediction Using Dynamic Bayesian Networks

PubMed Central

Reynolds, Sheila M.; Käll, Lukas; Riffle, Michael E.; Bilmes, Jeff A.; Noble, William Stafford

2008-01-01

Hidden Markov models (HMMs) have been successfully applied to the tasks of transmembrane protein topology prediction and signal peptide prediction. In this paper we expand upon this work by making use of the more powerful class of dynamic Bayesian networks (DBNs). Our model, Philius, is inspired by a previously published HMM, Phobius, and combines a signal peptide submodel with a transmembrane submodel. We introduce a two-stage DBN decoder that combines the power of posterior decoding with the grammar constraints of Viterbi-style decoding. Philius also provides protein type, segment, and topology confidence metrics to aid in the interpretation of the predictions. We report a relative improvement of 13% over Phobius in full-topology prediction accuracy on transmembrane proteins, and a sensitivity and specificity of 0.96 in detecting signal peptides. We also show that our confidence metrics correlate well with the observed precision. In addition, we have made predictions on all 6.3 million proteins in the Yeast Resource Center (YRC) database. This large-scale study provides an overall picture of the relative numbers of proteins that include a signal-peptide and/or one or more transmembrane segments as well as a valuable resource for the scientific community. All DBNs are implemented using the Graphical Models Toolkit. Source code for the models described here is available at http://noble.gs.washington.edu/proj/philius. A Philius Web server is available at http://www.yeastrc.org/philius, and the predictions on the YRC database are available at http://www.yeastrc.org/pdr. PMID:18989393

A Hierarchical Modeling Approach to Data Analysis and Study Design in a Multi-Site Experimental fMRI Study

ERIC Educational Resources Information Center

Zhou, Bo; Konstorum, Anna; Duong, Thao; Tieu, Kinh H.; Wells, William M.; Brown, Gregory G.; Stern, Hal S.; Shahbaba, Babak

2013-01-01

We propose a hierarchical Bayesian model for analyzing multi-site experimental fMRI studies. Our method takes the hierarchical structure of the data (subjects are nested within sites, and there are multiple observations per subject) into account and allows for modeling between-site variation. Using posterior predictive model checking and model…

Variational dynamic background model for keyword spotting in handwritten documents

NASA Astrophysics Data System (ADS)

Kumar, Gaurav; Wshah, Safwan; Govindaraju, Venu

2013-12-01

We propose a bayesian framework for keyword spotting in handwritten documents. This work is an extension to our previous work where we proposed dynamic background model, DBM for keyword spotting that takes into account the local character level scores and global word level scores to learn a logistic regression classifier to separate keywords from non-keywords. In this work, we add a bayesian layer on top of the DBM called the variational dynamic background model, VDBM. The logistic regression classifier uses the sigmoid function to separate keywords from non-keywords. The sigmoid function being neither convex nor concave, exact inference of VDBM becomes intractable. An expectation maximization step is proposed to do approximate inference. The advantage of VDBM over the DBM is multi-fold. Firstly, being bayesian, it prevents over-fitting of data. Secondly, it provides better modeling of data and an improved prediction of unseen data. VDBM is evaluated on the IAM dataset and the results prove that it outperforms our prior work and other state of the art line based word spotting system.

Using robust Bayesian network to estimate the residuals of fluoroquinolone antibiotic in soil.

PubMed

Li, Xuewen; Xie, Yunfeng; Li, Lianfa; Yang, Xunfeng; Wang, Ning; Wang, Jinfeng

2015-11-01

Prediction of antibiotic pollution and its consequences is difficult, due to the uncertainties and complexities associated with multiple related factors. This article employed domain knowledge and spatial data to construct a Bayesian network (BN) model to assess fluoroquinolone antibiotic (FQs) pollution in the soil of an intensive vegetable cultivation area. The results show: (1) The relationships between FQs pollution and contributory factors: Three factors (cultivation methods, crop rotations, and chicken manure types) were consistently identified as predictors in the topological structures of three FQs, indicating their importance in FQs pollution; deduced with domain knowledge, the cultivation methods are determined by the crop rotations, which require different nutrients (derived from the manure) according to different plant biomass. (2) The performance of BN model: The integrative robust Bayesian network model achieved the highest detection probability (pd) of high-risk and receiver operating characteristic (ROC) area, since it incorporates domain knowledge and model uncertainty. Our encouraging findings have implications for the use of BN as a robust approach to assessment of FQs pollution and for informing decisions on appropriate remedial measures.

Bayesian Inference for Source Reconstruction: A Real-World Application

PubMed Central

Yee, Eugene; Hoffman, Ian; Ungar, Kurt

2014-01-01

This paper applies a Bayesian probabilistic inferential methodology for the reconstruction of the location and emission rate from an actual contaminant source (emission from the Chalk River Laboratories medical isotope production facility) using a small number of activity concentration measurements of a noble gas (Xenon-133) obtained from three stations that form part of the International Monitoring System radionuclide network. The sampling of the resulting posterior distribution of the source parameters is undertaken using a very efficient Markov chain Monte Carlo technique that utilizes a multiple-try differential evolution adaptive Metropolis algorithm with an archive of past states. It is shown that the principal difficulty in the reconstruction lay in the correct specification of the model errors (both scale and structure) for use in the Bayesian inferential methodology. In this context, two different measurement models for incorporation of the model error of the predicted concentrations are considered. The performance of both of these measurement models with respect to their accuracy and precision in the recovery of the source parameters is compared and contrasted. PMID:27379292

Bayesian modeling of Clostridium perfringens growth in beef-in-sauce products.

PubMed

Jaloustre, S; Cornu, M; Morelli, E; Noël, V; Delignette-Muller, M L

2011-04-01

Models on Clostridium perfringens growth which have been published to date have all been deterministic. A probabilistic model describing growth under non-isothermal conditions was thus proposed for predicting C. perfringens growth in beef-in-sauce products cooked and distributed in a French hospital. Model parameters were estimated from different types of data from various studies. A Bayesian approach was proposed to model the overall uncertainty regarding parameters and potential variability on the 'work to be done' (h(0)) during the germination, outgrowth and lag phase. Three models which differed according to their description of this parameter h(0) were tested. The model with inter-curve variability on h(0) was found to be the best one, on the basis of goodness-of-fit assessment and validation with literature data on results obtained under non-isothermal conditions. This model was used in two-dimensional Monte Carlo simulations to predict C. perfringens growth throughout the preparation of beef-in-sauce products, using temperature profiles recorded in a hospital kitchen. The median predicted growth was 7.8×10(-2) log(10) cfu·g(-1) (95% credibility interval [2.4×10(-2), 0.8]) despite the fact that for more than 50% of the registered temperature profiles cooling steps were longer than those required by French regulations. Copyright © 2010 Elsevier Ltd. All rights reserved.

A Bayesian Hierarchical Modeling Approach to Predicting Flow in Ungauged Basins

NASA Astrophysics Data System (ADS)

Gronewold, A.; Alameddine, I.; Anderson, R. M.

2009-12-01

Recent innovative approaches to identifying and applying regression-based relationships between land use patterns (such as increasing impervious surface area and decreasing vegetative cover) and rainfall-runoff model parameters represent novel and promising improvements to predicting flow from ungauged basins. In particular, these approaches allow for predicting flows under uncertain and potentially variable future conditions due to rapid land cover changes, variable climate conditions, and other factors. Despite the broad range of literature on estimating rainfall-runoff model parameters, however, the absence of a robust set of modeling tools for identifying and quantifying uncertainties in (and correlation between) rainfall-runoff model parameters represents a significant gap in current hydrological modeling research. Here, we build upon a series of recent publications promoting novel Bayesian and probabilistic modeling strategies for quantifying rainfall-runoff model parameter estimation uncertainty. Our approach applies alternative measures of rainfall-runoff model parameter joint likelihood (including Nash-Sutcliffe efficiency, among others) to simulate samples from the joint parameter posterior probability density function. We then use these correlated samples as response variables in a Bayesian hierarchical model with land use coverage data as predictor variables in order to develop a robust land use-based tool for forecasting flow in ungauged basins while accounting for, and explicitly acknowledging, parameter estimation uncertainty. We apply this modeling strategy to low-relief coastal watersheds of Eastern North Carolina, an area representative of coastal resource waters throughout the world because of its sensitive embayments and because of the abundant (but currently threatened) natural resources it hosts. Consequently, this area is the subject of several ongoing studies and large-scale planning initiatives, including those conducted through the United States Environmental Protection Agency (USEPA) total maximum daily load (TMDL) program, as well as those addressing coastal population dynamics and sea level rise. Our approach has several advantages, including the propagation of parameter uncertainty through a nonparametric probability distribution which avoids common pitfalls of fitting parameters and model error structure to a predetermined parametric distribution function. In addition, by explicitly acknowledging correlation between model parameters (and reflecting those correlations in our predictive model) our model yields relatively efficient prediction intervals (unlike those in the current literature which are often unnecessarily large, and may lead to overly-conservative management actions). Finally, our model helps improve understanding of the rainfall-runoff process by identifying model parameters (and associated catchment attributes) which are most sensitive to current and future land use change patterns. Disclaimer: Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.

Rapid recipe formulation for plasma etching of new materials

NASA Astrophysics Data System (ADS)

Chopra, Meghali; Zhang, Zizhuo; Ekerdt, John; Bonnecaze, Roger T.

2016-03-01

A fast and inexpensive scheme for etch rate prediction using flexible continuum models and Bayesian statistics is demonstrated. Bulk etch rates of MgO are predicted using a steady-state model with volume-averaged plasma parameters and classical Langmuir surface kinetics. Plasma particle and surface kinetics are modeled within a global plasma framework using single component Metropolis Hastings methods and limited data. The accuracy of these predictions is evaluated with synthetic and experimental etch rate data for magnesium oxide in an ICP-RIE system. This approach is compared and superior to factorial models generated from JMP, a software package frequently employed for recipe creation and optimization.

APPLICATION OF BAYESIAN AND GEOSTATISTICAL MODELING TO THE ENVIRONMENTAL MONITORING OF CS-137 AT THE IDAHO NATIONAL LABORATORY

DOE Office of Scientific and Technical Information (OSTI.GOV)

Kara G. Eby

2010-08-01

At the Idaho National Laboratory (INL) Cs-137 concentrations above the U.S. Environmental Protection Agency risk-based threshold of 0.23 pCi/g may increase the risk of human mortality due to cancer. As a leader in nuclear research, the INL has been conducting nuclear activities for decades. Elevated anthropogenic radionuclide levels including Cs-137 are a result of atmospheric weapons testing, the Chernobyl accident, and nuclear activities occurring at the INL site. Therefore environmental monitoring and long-term surveillance of Cs-137 is required to evaluate risk. However, due to the large land area involved, frequent and comprehensive monitoring is limited. Developing a spatial model thatmore » predicts Cs-137 concentrations at unsampled locations will enhance the spatial characterization of Cs-137 in surface soils, provide guidance for an efficient monitoring program, and pinpoint areas requiring mitigation strategies. The predictive model presented herein is based on applied geostatistics using a Bayesian analysis of environmental characteristics across the INL site, which provides kriging spatial maps of both Cs-137 estimates and prediction errors. Comparisons are presented of two different kriging methods, showing that the use of secondary information (i.e., environmental characteristics) can provide improved prediction performance in some areas of the INL site.« less

Coupling of Bayesian Networks with GIS for wildfire risk assessment on natural and agricultural areas of the Mediterranean

NASA Astrophysics Data System (ADS)

Scherb, Anke; Papakosta, Panagiota; Straub, Daniel

2014-05-01

Wildfires cause severe damages to ecosystems, socio-economic assets, and human lives in the Mediterranean. To facilitate coping with wildfire risks, an understanding of the factors influencing wildfire occurrence and behavior (e.g. human activity, weather conditions, topography, fuel loads) and their interaction is of importance, as is the implementation of this knowledge in improved wildfire hazard and risk prediction systems. In this project, a probabilistic wildfire risk prediction model is developed, with integrated fire occurrence and fire propagation probability and potential impact prediction on natural and cultivated areas. Bayesian Networks (BNs) are used to facilitate the probabilistic modeling. The final BN model is a spatial-temporal prediction system at the meso scale (1 km2 spatial and 1 day temporal resolution). The modeled consequences account for potential restoration costs and production losses referred to forests, agriculture, and (semi-) natural areas. BNs and a geographic information system (GIS) are coupled within this project to support a semi-automated BN model parameter learning and the spatial-temporal risk prediction. The coupling also enables the visualization of prediction results by means of daily maps. The BN parameters are learnt for Cyprus with data from 2006-2009. Data from 2010 is used as validation data set. A special focus is put on the performance evaluation of the BN for fire occurrence, which is modeled as binary classifier and thus, could be validated by means of Receiver Operator Characteristic (ROC) curves. With the final best models, AUC values of more than 70% for validation could be achieved, which indicates potential for reliable prediction performance via BN. Maps of selected days in 2010 are shown to illustrate final prediction results. The resulting system can be easily expanded to predict additional expected damages in the mesoscale (e.g. building and infrastructure damages). The system can support planning of preventive measures (e.g. state resources allocation for wildfire prevention and preparedness) and assist recuperation plans of damaged areas.

Predicting the heat of vaporization of iron at high temperatures using time-resolved laser-induced incandescence and Bayesian model selection

NASA Astrophysics Data System (ADS)

Sipkens, Timothy A.; Hadwin, Paul J.; Grauer, Samuel J.; Daun, Kyle J.

2018-03-01

Competing theories have been proposed to account for how the latent heat of vaporization of liquid iron varies with temperature, but experimental confirmation remains elusive, particularly at high temperatures. We propose time-resolved laser-induced incandescence measurements on iron nanoparticles combined with Bayesian model plausibility, as a novel method for evaluating these relationships. Our approach scores the explanatory power of candidate models, accounting for parameter uncertainty, model complexity, measurement noise, and goodness-of-fit. The approach is first validated with simulated data and then applied to experimental data for iron nanoparticles in argon. Our results justify the use of Román's equation to account for the temperature dependence of the latent heat of vaporization of liquid iron.

Bayesian model averaging using particle filtering and Gaussian mixture modeling: Theory, concepts, and simulation experiments

NASA Astrophysics Data System (ADS)

Rings, Joerg; Vrugt, Jasper A.; Schoups, Gerrit; Huisman, Johan A.; Vereecken, Harry

2012-05-01

Bayesian model averaging (BMA) is a standard method for combining predictive distributions from different models. In recent years, this method has enjoyed widespread application and use in many fields of study to improve the spread-skill relationship of forecast ensembles. The BMA predictive probability density function (pdf) of any quantity of interest is a weighted average of pdfs centered around the individual (possibly bias-corrected) forecasts, where the weights are equal to posterior probabilities of the models generating the forecasts, and reflect the individual models skill over a training (calibration) period. The original BMA approach presented by Raftery et al. (2005) assumes that the conditional pdf of each individual model is adequately described with a rather standard Gaussian or Gamma statistical distribution, possibly with a heteroscedastic variance. Here we analyze the advantages of using BMA with a flexible representation of the conditional pdf. A joint particle filtering and Gaussian mixture modeling framework is presented to derive analytically, as closely and consistently as possible, the evolving forecast density (conditional pdf) of each constituent ensemble member. The median forecasts and evolving conditional pdfs of the constituent models are subsequently combined using BMA to derive one overall predictive distribution. This paper introduces the theory and concepts of this new ensemble postprocessing method, and demonstrates its usefulness and applicability by numerical simulation of the rainfall-runoff transformation using discharge data from three different catchments in the contiguous United States. The revised BMA method receives significantly lower-prediction errors than the original default BMA method (due to filtering) with predictive uncertainty intervals that are substantially smaller but still statistically coherent (due to the use of a time-variant conditional pdf).

Predicting the multi-domain progression of Parkinson's disease: a Bayesian multivariate generalized linear mixed-effect model.

PubMed

Wang, Ming; Li, Zheng; Lee, Eun Young; Lewis, Mechelle M; Zhang, Lijun; Sterling, Nicholas W; Wagner, Daymond; Eslinger, Paul; Du, Guangwei; Huang, Xuemei

2017-09-25

It is challenging for current statistical models to predict clinical progression of Parkinson's disease (PD) because of the involvement of multi-domains and longitudinal data. Past univariate longitudinal or multivariate analyses from cross-sectional trials have limited power to predict individual outcomes or a single moment. The multivariate generalized linear mixed-effect model (GLMM) under the Bayesian framework was proposed to study multi-domain longitudinal outcomes obtained at baseline, 18-, and 36-month. The outcomes included motor, non-motor, and postural instability scores from the MDS-UPDRS, and demographic and standardized clinical data were utilized as covariates. The dynamic prediction was performed for both internal and external subjects using the samples from the posterior distributions of the parameter estimates and random effects, and also the predictive accuracy was evaluated based on the root of mean square error (RMSE), absolute bias (AB) and the area under the receiver operating characteristic (ROC) curve. First, our prediction model identified clinical data that were differentially associated with motor, non-motor, and postural stability scores. Second, the predictive accuracy of our model for the training data was assessed, and improved prediction was gained in particularly for non-motor (RMSE and AB: 2.89 and 2.20) compared to univariate analysis (RMSE and AB: 3.04 and 2.35). Third, the individual-level predictions of longitudinal trajectories for the testing data were performed, with ~80% observed values falling within the 95% credible intervals. Multivariate general mixed models hold promise to predict clinical progression of individual outcomes in PD. The data was obtained from Dr. Xuemei Huang's NIH grant R01 NS060722 , part of NINDS PD Biomarker Program (PDBP). All data was entered within 24 h of collection to the Data Management Repository (DMR), which is publically available ( https://pdbp.ninds.nih.gov/data-management ).

On the Bayesian Treed Multivariate Gaussian Process with Linear Model of Coregionalization

DOE Office of Scientific and Technical Information (OSTI.GOV)

Konomi, Bledar A.; Karagiannis, Georgios; Lin, Guang

2015-02-01

The Bayesian treed Gaussian process (BTGP) has gained popularity in recent years because it provides a straightforward mechanism for modeling non-stationary data and can alleviate computational demands by fitting models to less data. The extension of BTGP to the multivariate setting requires us to model the cross-covariance and to propose efficient algorithms that can deal with trans-dimensional MCMC moves. In this paper we extend the cross-covariance of the Bayesian treed multivariate Gaussian process (BTMGP) to that of linear model of Coregionalization (LMC) cross-covariances. Different strategies have been developed to improve the MCMC mixing and invert smaller matrices in the Bayesianmore » inference. Moreover, we compare the proposed BTMGP with existing multiple BTGP and BTMGP in test cases and multiphase flow computer experiment in a full scale regenerator of a carbon capture unit. The use of the BTMGP with LMC cross-covariance helped to predict the computer experiments relatively better than existing competitors. The proposed model has a wide variety of applications, such as computer experiments and environmental data. In the case of computer experiments we also develop an adaptive sampling strategy for the BTMGP with LMC cross-covariance function.« less

Poisson Mixture Regression Models for Heart Disease Prediction.

PubMed

Mufudza, Chipo; Erol, Hamza

2016-01-01

Early heart disease control can be achieved by high disease prediction and diagnosis efficiency. This paper focuses on the use of model based clustering techniques to predict and diagnose heart disease via Poisson mixture regression models. Analysis and application of Poisson mixture regression models is here addressed under two different classes: standard and concomitant variable mixture regression models. Results show that a two-component concomitant variable Poisson mixture regression model predicts heart disease better than both the standard Poisson mixture regression model and the ordinary general linear Poisson regression model due to its low Bayesian Information Criteria value. Furthermore, a Zero Inflated Poisson Mixture Regression model turned out to be the best model for heart prediction over all models as it both clusters individuals into high or low risk category and predicts rate to heart disease componentwise given clusters available. It is deduced that heart disease prediction can be effectively done by identifying the major risks componentwise using Poisson mixture regression model.

Poisson Mixture Regression Models for Heart Disease Prediction

PubMed Central

Erol, Hamza

2016-01-01

Early heart disease control can be achieved by high disease prediction and diagnosis efficiency. This paper focuses on the use of model based clustering techniques to predict and diagnose heart disease via Poisson mixture regression models. Analysis and application of Poisson mixture regression models is here addressed under two different classes: standard and concomitant variable mixture regression models. Results show that a two-component concomitant variable Poisson mixture regression model predicts heart disease better than both the standard Poisson mixture regression model and the ordinary general linear Poisson regression model due to its low Bayesian Information Criteria value. Furthermore, a Zero Inflated Poisson Mixture Regression model turned out to be the best model for heart prediction over all models as it both clusters individuals into high or low risk category and predicts rate to heart disease componentwise given clusters available. It is deduced that heart disease prediction can be effectively done by identifying the major risks componentwise using Poisson mixture regression model. PMID:27999611

Bayesian just-so stories in psychology and neuroscience.

PubMed

Bowers, Jeffrey S; Davis, Colin J

2012-05-01

According to Bayesian theories in psychology and neuroscience, minds and brains are (near) optimal in solving a wide range of tasks. We challenge this view and argue that more traditional, non-Bayesian approaches are more promising. We make 3 main arguments. First, we show that the empirical evidence for Bayesian theories in psychology is weak. This weakness relates to the many arbitrary ways that priors, likelihoods, and utility functions can be altered in order to account for the data that are obtained, making the models unfalsifiable. It further relates to the fact that Bayesian theories are rarely better at predicting data compared with alternative (and simpler) non-Bayesian theories. Second, we show that the empirical evidence for Bayesian theories in neuroscience is weaker still. There are impressive mathematical analyses showing how populations of neurons could compute in a Bayesian manner but little or no evidence that they do. Third, we challenge the general scientific approach that characterizes Bayesian theorizing in cognitive science. A common premise is that theories in psychology should largely be constrained by a rational analysis of what the mind ought to do. We question this claim and argue that many of the important constraints come from biological, evolutionary, and processing (algorithmic) considerations that have no adaptive relevance to the problem per se. In our view, these factors have contributed to the development of many Bayesian "just so" stories in psychology and neuroscience; that is, mathematical analyses of cognition that can be used to explain almost any behavior as optimal. 2012 APA, all rights reserved.

Uncertainty estimation of a complex water quality model: The influence of Box-Cox transformation on Bayesian approaches and comparison with a non-Bayesian method

NASA Astrophysics Data System (ADS)

Freni, Gabriele; Mannina, Giorgio

In urban drainage modelling, uncertainty analysis is of undoubted necessity. However, uncertainty analysis in urban water-quality modelling is still in its infancy and only few studies have been carried out. Therefore, several methodological aspects still need to be experienced and clarified especially regarding water quality modelling. The use of the Bayesian approach for uncertainty analysis has been stimulated by its rigorous theoretical framework and by the possibility of evaluating the impact of new knowledge on the modelling predictions. Nevertheless, the Bayesian approach relies on some restrictive hypotheses that are not present in less formal methods like the Generalised Likelihood Uncertainty Estimation (GLUE). One crucial point in the application of Bayesian method is the formulation of a likelihood function that is conditioned by the hypotheses made regarding model residuals. Statistical transformations, such as the use of Box-Cox equation, are generally used to ensure the homoscedasticity of residuals. However, this practice may affect the reliability of the analysis leading to a wrong uncertainty estimation. The present paper aims to explore the influence of the Box-Cox equation for environmental water quality models. To this end, five cases were considered one of which was the “real” residuals distributions (i.e. drawn from available data). The analysis was applied to the Nocella experimental catchment (Italy) which is an agricultural and semi-urbanised basin where two sewer systems, two wastewater treatment plants and a river reach were monitored during both dry and wet weather periods. The results show that the uncertainty estimation is greatly affected by residual transformation and a wrong assumption may also affect the evaluation of model uncertainty. The use of less formal methods always provide an overestimation of modelling uncertainty with respect to Bayesian method but such effect is reduced if a wrong assumption is made regarding the residuals distribution. If residuals are not normally distributed, the uncertainty is over-estimated if Box-Cox transformation is not applied or non-calibrated parameter is used.

Attention in the predictive mind.

PubMed

Ransom, Madeleine; Fazelpour, Sina; Mole, Christopher

2017-01-01

It has recently become popular to suggest that cognition can be explained as a process of Bayesian prediction error minimization. Some advocates of this view propose that attention should be understood as the optimization of expected precisions in the prediction-error signal (Clark, 2013, 2016; Feldman & Friston, 2010; Hohwy, 2012, 2013). This proposal successfully accounts for several attention-related phenomena. We claim that it cannot account for all of them, since there are certain forms of voluntary attention that it cannot accommodate. We therefore suggest that, although the theory of Bayesian prediction error minimization introduces some powerful tools for the explanation of mental phenomena, its advocates have been wrong to claim that Bayesian prediction error minimization is 'all the brain ever does'. Copyright © 2016 Elsevier Inc. All rights reserved.

Taming Many-Parameter BSM Models with Bayesian Neural Networks

NASA Astrophysics Data System (ADS)

Kuchera, M. P.; Karbo, A.; Prosper, H. B.; Sanchez, A.; Taylor, J. Z.

2017-09-01

The search for physics Beyond the Standard Model (BSM) is a major focus of large-scale high energy physics experiments. One method is to look for specific deviations from the Standard Model that are predicted by BSM models. In cases where the model has a large number of free parameters, standard search methods become intractable due to computation time. This talk presents results using Bayesian Neural Networks, a supervised machine learning method, to enable the study of higher-dimensional models. The popular phenomenological Minimal Supersymmetric Standard Model was studied as an example of the feasibility and usefulness of this method. Graphics Processing Units (GPUs) are used to expedite the calculations. Cross-section predictions for 13 TeV proton collisions will be presented. My participation in the Conference Experience for Undergraduates (CEU) in 2004-2006 exposed me to the national and global significance of cutting-edge research. At the 2005 CEU, I presented work from the previous summer's SULI internship at Lawrence Berkeley Laboratory, where I learned to program while working on the Majorana Project. That work inspired me to follow a similar research path, which led me to my current work on computational methods applied to BSM physics.

Bayesian Design of Superiority Clinical Trials for Recurrent Events Data with Applications to Bleeding and Transfusion Events in Myelodyplastic Syndrome

PubMed Central

Chen, Ming-Hui; Zeng, Donglin; Hu, Kuolung; Jia, Catherine

2014-01-01

Summary In many biomedical studies, patients may experience the same type of recurrent event repeatedly over time, such as bleeding, multiple infections and disease. In this article, we propose a Bayesian design to a pivotal clinical trial in which lower risk myelodysplastic syndromes (MDS) patients are treated with MDS disease modifying therapies. One of the key study objectives is to demonstrate the investigational product (treatment) effect on reduction of platelet transfusion and bleeding events while receiving MDS therapies. In this context, we propose a new Bayesian approach for the design of superiority clinical trials using recurrent events frailty regression models. Historical recurrent events data from an already completed phase 2 trial are incorporated into the Bayesian design via the partial borrowing power prior of Ibrahim et al. (2012, Biometrics 68, 578–586). An efficient Gibbs sampling algorithm, a predictive data generation algorithm, and a simulation-based algorithm are developed for sampling from the fitting posterior distribution, generating the predictive recurrent events data, and computing various design quantities such as the type I error rate and power, respectively. An extensive simulation study is conducted to compare the proposed method to the existing frequentist methods and to investigate various operating characteristics of the proposed design. PMID:25041037

Multimodel Ensemble Methods for Prediction of Wake-Vortex Transport and Decay Originating NASA

NASA Technical Reports Server (NTRS)

Korner, Stephan; Ahmad, Nashat N.; Holzapfel, Frank; VanValkenburg, Randal L.

2017-01-01

Several multimodel ensemble methods are selected and further developed to improve the deterministic and probabilistic prediction skills of individual wake-vortex transport and decay models. The different multimodel ensemble methods are introduced, and their suitability for wake applications is demonstrated. The selected methods include direct ensemble averaging, Bayesian model averaging, and Monte Carlo simulation. The different methodologies are evaluated employing data from wake-vortex field measurement campaigns conducted in the United States and Germany.

A Bayesian approach to estimating hidden variables as well as missing and wrong molecular interactions in ordinary differential equation-based mathematical models.

PubMed

Engelhardt, Benjamin; Kschischo, Maik; Fröhlich, Holger

2017-06-01

Ordinary differential equations (ODEs) are a popular approach to quantitatively model molecular networks based on biological knowledge. However, such knowledge is typically restricted. Wrongly modelled biological mechanisms as well as relevant external influence factors that are not included into the model are likely to manifest in major discrepancies between model predictions and experimental data. Finding the exact reasons for such observed discrepancies can be quite challenging in practice. In order to address this issue, we suggest a Bayesian approach to estimate hidden influences in ODE-based models. The method can distinguish between exogenous and endogenous hidden influences. Thus, we can detect wrongly specified as well as missed molecular interactions in the model. We demonstrate the performance of our Bayesian dynamic elastic-net with several ordinary differential equation models from the literature, such as human JAK-STAT signalling, information processing at the erythropoietin receptor, isomerization of liquid α -Pinene, G protein cycling in yeast and UV-B triggered signalling in plants. Moreover, we investigate a set of commonly known network motifs and a gene-regulatory network. Altogether our method supports the modeller in an algorithmic manner to identify possible sources of errors in ODE-based models on the basis of experimental data. © 2017 The Author(s).

Convergence analysis of surrogate-based methods for Bayesian inverse problems

NASA Astrophysics Data System (ADS)

Yan, Liang; Zhang, Yuan-Xiang

2017-12-01

The major challenges in the Bayesian inverse problems arise from the need for repeated evaluations of the forward model, as required by Markov chain Monte Carlo (MCMC) methods for posterior sampling. Many attempts at accelerating Bayesian inference have relied on surrogates for the forward model, typically constructed through repeated forward simulations that are performed in an offline phase. Although such approaches can be quite effective at reducing computation cost, there has been little analysis of the approximation on posterior inference. In this work, we prove error bounds on the Kullback-Leibler (KL) distance between the true posterior distribution and the approximation based on surrogate models. Our rigorous error analysis show that if the forward model approximation converges at certain rate in the prior-weighted L 2 norm, then the posterior distribution generated by the approximation converges to the true posterior at least two times faster in the KL sense. The error bound on the Hellinger distance is also provided. To provide concrete examples focusing on the use of the surrogate model based methods, we present an efficient technique for constructing stochastic surrogate models to accelerate the Bayesian inference approach. The Christoffel least squares algorithms, based on generalized polynomial chaos, are used to construct a polynomial approximation of the forward solution over the support of the prior distribution. The numerical strategy and the predicted convergence rates are then demonstrated on the nonlinear inverse problems, involving the inference of parameters appearing in partial differential equations.

Nonlinear-drifted Brownian motion with multiple hidden states for remaining useful life prediction of rechargeable batteries

NASA Astrophysics Data System (ADS)

Wang, Dong; Zhao, Yang; Yang, Fangfang; Tsui, Kwok-Leung

2017-09-01

Brownian motion with adaptive drift has attracted much attention in prognostics because its first hitting time is highly relevant to remaining useful life prediction and it follows the inverse Gaussian distribution. Besides linear degradation modeling, nonlinear-drifted Brownian motion has been developed to model nonlinear degradation. Moreover, the first hitting time distribution of the nonlinear-drifted Brownian motion has been approximated by time-space transformation. In the previous studies, the drift coefficient is the only hidden state used in state space modeling of the nonlinear-drifted Brownian motion. Besides the drift coefficient, parameters of a nonlinear function used in the nonlinear-drifted Brownian motion should be treated as additional hidden states of state space modeling to make the nonlinear-drifted Brownian motion more flexible. In this paper, a prognostic method based on nonlinear-drifted Brownian motion with multiple hidden states is proposed and then it is applied to predict remaining useful life of rechargeable batteries. 26 sets of rechargeable battery degradation samples are analyzed to validate the effectiveness of the proposed prognostic method. Moreover, some comparisons with a standard particle filter based prognostic method, a spherical cubature particle filter based prognostic method and two classic Bayesian prognostic methods are conducted to highlight the superiority of the proposed prognostic method. Results show that the proposed prognostic method has lower average prediction errors than the particle filter based prognostic methods and the classic Bayesian prognostic methods for battery remaining useful life prediction.

Novelty and Inductive Generalization in Human Reinforcement Learning.

PubMed

Gershman, Samuel J; Niv, Yael

2015-07-01

In reinforcement learning (RL), a decision maker searching for the most rewarding option is often faced with the question: What is the value of an option that has never been tried before? One way to frame this question is as an inductive problem: How can I generalize my previous experience with one set of options to a novel option? We show how hierarchical Bayesian inference can be used to solve this problem, and we describe an equivalence between the Bayesian model and temporal difference learning algorithms that have been proposed as models of RL in humans and animals. According to our view, the search for the best option is guided by abstract knowledge about the relationships between different options in an environment, resulting in greater search efficiency compared to traditional RL algorithms previously applied to human cognition. In two behavioral experiments, we test several predictions of our model, providing evidence that humans learn and exploit structured inductive knowledge to make predictions about novel options. In light of this model, we suggest a new interpretation of dopaminergic responses to novelty. Copyright © 2015 Cognitive Science Society, Inc.

Novelty and Inductive Generalization in Human Reinforcement Learning

PubMed Central

Gershman, Samuel J.; Niv, Yael

2015-01-01

In reinforcement learning, a decision maker searching for the most rewarding option is often faced with the question: what is the value of an option that has never been tried before? One way to frame this question is as an inductive problem: how can I generalize my previous experience with one set of options to a novel option? We show how hierarchical Bayesian inference can be used to solve this problem, and describe an equivalence between the Bayesian model and temporal difference learning algorithms that have been proposed as models of reinforcement learning in humans and animals. According to our view, the search for the best option is guided by abstract knowledge about the relationships between different options in an environment, resulting in greater search efficiency compared to traditional reinforcement learning algorithms previously applied to human cognition. In two behavioral experiments, we test several predictions of our model, providing evidence that humans learn and exploit structured inductive knowledge to make predictions about novel options. In light of this model, we suggest a new interpretation of dopaminergic responses to novelty. PMID:25808176

Predicting motor vehicle collisions using Bayesian neural network models: an empirical analysis.

PubMed

Xie, Yuanchang; Lord, Dominique; Zhang, Yunlong

2007-09-01

Statistical models have frequently been used in highway safety studies. They can be utilized for various purposes, including establishing relationships between variables, screening covariates and predicting values. Generalized linear models (GLM) and hierarchical Bayes models (HBM) have been the most common types of model favored by transportation safety analysts. Over the last few years, researchers have proposed the back-propagation neural network (BPNN) model for modeling the phenomenon under study. Compared to GLMs and HBMs, BPNNs have received much less attention in highway safety modeling. The reasons are attributed to the complexity for estimating this kind of model as well as the problem related to "over-fitting" the data. To circumvent the latter problem, some statisticians have proposed the use of Bayesian neural network (BNN) models. These models have been shown to perform better than BPNN models while at the same time reducing the difficulty associated with over-fitting the data. The objective of this study is to evaluate the application of BNN models for predicting motor vehicle crashes. To accomplish this objective, a series of models was estimated using data collected on rural frontage roads in Texas. Three types of models were compared: BPNN, BNN and the negative binomial (NB) regression models. The results of this study show that in general both types of neural network models perform better than the NB regression model in terms of data prediction. Although the BPNN model can occasionally provide better or approximately equivalent prediction performance compared to the BNN model, in most cases its prediction performance is worse than the BNN model. In addition, the data fitting performance of the BPNN model is consistently worse than the BNN model, which suggests that the BNN model has better generalization abilities than the BPNN model and can effectively alleviate the over-fitting problem without significantly compromising the nonlinear approximation ability. The results also show that BNNs could be used for other useful analyses in highway safety, including the development of accident modification factors and for improving the prediction capabilities for evaluating different highway design alternatives.

Maximum entropy perception-action space: a Bayesian model of eye movement selection

NASA Astrophysics Data System (ADS)

Colas, Francis; Bessière, Pierre; Girard, Benoît

2011-03-01

In this article, we investigate the issue of the selection of eye movements in a free-eye Multiple Object Tracking task. We propose a Bayesian model of retinotopic maps with a complex logarithmic mapping. This model is structured in two parts: a representation of the visual scene, and a decision model based on the representation. We compare different decision models based on different features of the representation and we show that taking into account uncertainty helps predict the eye movements of subjects recorded in a psychophysics experiment. Finally, based on experimental data, we postulate that the complex logarithmic mapping has a functional relevance, as the density of objects in this space in more uniform than expected. This may indicate that the representation space and control strategies are such that the object density is of maximum entropy.

A Bayesian-based system to assess wave-driven flooding hazards on coral reef-lined coasts

USGS Publications Warehouse

Pearson, S. G.; Storlazzi, Curt; van Dongeren, A. R.; Tissier, M. F. S.; Reniers, A. J. H. M.

2017-01-01

Many low-elevation, coral reef-lined, tropical coasts are vulnerable to the effects of climate change, sea level rise, and wave-induced flooding. The considerable morphological diversity of these coasts and the variability of the hydrodynamic forcing that they are exposed to make predicting wave-induced flooding a challenge. A process-based wave-resolving hydrodynamic model (XBeach Non-Hydrostatic, “XBNH”) was used to create a large synthetic database for use in a “Bayesian Estimator for Wave Attack in Reef Environments” (BEWARE), relating incident hydrodynamics and coral reef geomorphology to coastal flooding hazards on reef-lined coasts. Building on previous work, BEWARE improves system understanding of reef hydrodynamics by examining the intrinsic reef and extrinsic forcing factors controlling runup and flooding on reef-lined coasts. The Bayesian estimator has high predictive skill for the XBNH model outputs that are flooding indicators, and was validated for a number of available field cases. It was found that, in order to accurately predict flooding hazards, water depth over the reef flat, incident wave conditions, and reef flat width are the most essential factors, whereas other factors such as beach slope and bed friction due to the presence or absence of corals are less important. BEWARE is a potentially powerful tool for use in early warning systems or risk assessment studies, and can be used to make projections about how wave-induced flooding on coral reef-lined coasts may change due to climate change.

Impact of censoring on learning Bayesian networks in survival modelling.

PubMed

Stajduhar, Ivan; Dalbelo-Basić, Bojana; Bogunović, Nikola

2009-11-01

Bayesian networks are commonly used for presenting uncertainty and covariate interactions in an easily interpretable way. Because of their efficient inference and ability to represent causal relationships, they are an excellent choice for medical decision support systems in diagnosis, treatment, and prognosis. Although good procedures for learning Bayesian networks from data have been defined, their performance in learning from censored survival data has not been widely studied. In this paper, we explore how to use these procedures to learn about possible interactions between prognostic factors and their influence on the variate of interest. We study how censoring affects the probability of learning correct Bayesian network structures. Additionally, we analyse the potential usefulness of the learnt models for predicting the time-independent probability of an event of interest. We analysed the influence of censoring with a simulation on synthetic data sampled from randomly generated Bayesian networks. We used two well-known methods for learning Bayesian networks from data: a constraint-based method and a score-based method. We compared the performance of each method under different levels of censoring to those of the naive Bayes classifier and the proportional hazards model. We did additional experiments on several datasets from real-world medical domains. The machine-learning methods treated censored cases in the data as event-free. We report and compare results for several commonly used model evaluation metrics. On average, the proportional hazards method outperformed other methods in most censoring setups. As part of the simulation study, we also analysed structural similarities of the learnt networks. Heavy censoring, as opposed to no censoring, produces up to a 5% surplus and up to 10% missing total arcs. It also produces up to 50% missing arcs that should originally be connected to the variate of interest. Presented methods for learning Bayesian networks from data can be used to learn from censored survival data in the presence of light censoring (up to 20%) by treating censored cases as event-free. Given intermediate or heavy censoring, the learnt models become tuned to the majority class and would thus require a different approach.

Bayesian Decision Support for Adaptive Lung Treatments

NASA Astrophysics Data System (ADS)

McShan, Daniel; Luo, Yi; Schipper, Matt; TenHaken, Randall

2014-03-01

Purpose: A Bayesian Decision Network will be demonstrated to provide clinical decision support for adaptive lung response-driven treatment management based on evidence that physiologic metrics may correlate better with individual patient response than traditional (population-based) dose and volume-based metrics. Further, there is evidence that information obtained during the course of radiation therapy may further improve response predictions. Methods: Clinical factors were gathered for 58 patients including planned mean lung dose, and the bio-markers IL-8 and TGF-β1 obtained prior to treatment and two weeks into treatment along with complication outcomes for these patients. A Bayesian Decision Network was constructed using Netica 5.0.2 from Norsys linking these clinical factors to obtain a prediction of radiation induced lung disese (RILD) complication. A decision node was added to the network to provide a plan adaption recommendation based on the trade-off between the RILD prediction and complexity of replanning. A utility node provides the weighting cost between the competing factors. Results: The decision node predictions were optimized against the data for the 58 cases. With this decision network solution, one can consider the decision result for a new patient with specific findings to obtain a recommendation to adaptively modify the originally planned treatment course. Conclusions: A Bayesian approach allows handling and propagating probabilistic data in a logical and principled manner. Decision networks provide the further ability to provide utility-based trade-offs, reflecting non-medical but practical cost/benefit analysis. The network demonstrated illustrates the basic concept, but many other factors may affect these decisions and work on building better models are being designed and tested. Acknowledgement: Supported by NIH-P01-CA59827

Bayesian dynamic modeling of time series of dengue disease case counts.

PubMed

Martínez-Bello, Daniel Adyro; López-Quílez, Antonio; Torres-Prieto, Alexander

2017-07-01

The aim of this study is to model the association between weekly time series of dengue case counts and meteorological variables, in a high-incidence city of Colombia, applying Bayesian hierarchical dynamic generalized linear models over the period January 2008 to August 2015. Additionally, we evaluate the model's short-term performance for predicting dengue cases. The methodology shows dynamic Poisson log link models including constant or time-varying coefficients for the meteorological variables. Calendar effects were modeled using constant or first- or second-order random walk time-varying coefficients. The meteorological variables were modeled using constant coefficients and first-order random walk time-varying coefficients. We applied Markov Chain Monte Carlo simulations for parameter estimation, and deviance information criterion statistic (DIC) for model selection. We assessed the short-term predictive performance of the selected final model, at several time points within the study period using the mean absolute percentage error. The results showed the best model including first-order random walk time-varying coefficients for calendar trend and first-order random walk time-varying coefficients for the meteorological variables. Besides the computational challenges, interpreting the results implies a complete analysis of the time series of dengue with respect to the parameter estimates of the meteorological effects. We found small values of the mean absolute percentage errors at one or two weeks out-of-sample predictions for most prediction points, associated with low volatility periods in the dengue counts. We discuss the advantages and limitations of the dynamic Poisson models for studying the association between time series of dengue disease and meteorological variables. The key conclusion of the study is that dynamic Poisson models account for the dynamic nature of the variables involved in the modeling of time series of dengue disease, producing useful models for decision-making in public health.

A Bayesian framework based on a Gaussian mixture model and radial-basis-function Fisher discriminant analysis (BayGmmKda V1.1) for spatial prediction of floods

NASA Astrophysics Data System (ADS)

Tien Bui, Dieu; Hoang, Nhat-Duc

2017-09-01

In this study, a probabilistic model, named as BayGmmKda, is proposed for flood susceptibility assessment in a study area in central Vietnam. The new model is a Bayesian framework constructed by a combination of a Gaussian mixture model (GMM), radial-basis-function Fisher discriminant analysis (RBFDA), and a geographic information system (GIS) database. In the Bayesian framework, GMM is used for modeling the data distribution of flood-influencing factors in the GIS database, whereas RBFDA is utilized to construct a latent variable that aims at enhancing the model performance. As a result, the posterior probabilistic output of the BayGmmKda model is used as flood susceptibility index. Experiment results showed that the proposed hybrid framework is superior to other benchmark models, including the adaptive neuro-fuzzy inference system and the support vector machine. To facilitate the model implementation, a software program of BayGmmKda has been developed in MATLAB. The BayGmmKda program can accurately establish a flood susceptibility map for the study region. Accordingly, local authorities can overlay this susceptibility map onto various land-use maps for the purpose of land-use planning or management.

LiDAR based prediction of forest biomass using hierarchical models with spatially varying coefficients

Treesearch

Chad Babcock; Andrew O. Finley; John B. Bradford; Randy Kolka; Richard Birdsey; Michael G. Ryan

2015-01-01

Many studies and production inventory systems have shown the utility of coupling covariates derived from Light Detection and Ranging (LiDAR) data with forest variables measured on georeferenced inventory plots through regression models. The objective of this study was to propose and assess the use of a Bayesian hierarchical modeling framework that accommodates both...

Statistical Bayesian method for reliability evaluation based on ADT data

NASA Astrophysics Data System (ADS)

Lu, Dawei; Wang, Lizhi; Sun, Yusheng; Wang, Xiaohong

2018-05-01

Accelerated degradation testing (ADT) is frequently conducted in the laboratory to predict the products’ reliability under normal operating conditions. Two kinds of methods, degradation path models and stochastic process models, are utilized to analyze degradation data and the latter one is the most popular method. However, some limitations like imprecise solution process and estimation result of degradation ratio still exist, which may affect the accuracy of the acceleration model and the extrapolation value. Moreover, the conducted solution of this problem, Bayesian method, lose key information when unifying the degradation data. In this paper, a new data processing and parameter inference method based on Bayesian method is proposed to handle degradation data and solve the problems above. First, Wiener process and acceleration model is chosen; Second, the initial values of degradation model and parameters of prior and posterior distribution under each level is calculated with updating and iteration of estimation values; Third, the lifetime and reliability values are estimated on the basis of the estimation parameters; Finally, a case study is provided to demonstrate the validity of the proposed method. The results illustrate that the proposed method is quite effective and accuracy in estimating the lifetime and reliability of a product.

Development of an On-board Failure Diagnostics and Prognostics System for Solid Rocket Booster

NASA Technical Reports Server (NTRS)

Smelyanskiy, Vadim N.; Luchinsky, Dmitry G.; Osipov, Vyatcheslav V.; Timucin, Dogan A.; Uckun, Serdar

2009-01-01

We develop a case breach model for the on-board fault diagnostics and prognostics system for subscale solid-rocket boosters (SRBs). The model development was motivated by recent ground firing tests, in which a deviation of measured time-traces from the predicted time-series was observed. A modified model takes into account the nozzle ablation, including the effect of roughness of the nozzle surface, the geometry of the fault, and erosion and burning of the walls of the hole in the metal case. The derived low-dimensional performance model (LDPM) of the fault can reproduce the observed time-series data very well. To verify the performance of the LDPM we build a FLUENT model of the case breach fault and demonstrate a good agreement between theoretical predictions based on the analytical solution of the model equations and the results of the FLUENT simulations. We then incorporate the derived LDPM into an inferential Bayesian framework and verify performance of the Bayesian algorithm for the diagnostics and prognostics of the case breach fault. It is shown that the obtained LDPM allows one to track parameters of the SRB during the flight in real time, to diagnose case breach fault, and to predict its values in the future. The application of the method to fault diagnostics and prognostics (FD&P) of other SRB faults modes is discussed.

Bayesian coronal seismology

NASA Astrophysics Data System (ADS)

Arregui, Iñigo

2018-01-01

In contrast to the situation in a laboratory, the study of the solar atmosphere has to be pursued without direct access to the physical conditions of interest. Information is therefore incomplete and uncertain and inference methods need to be employed to diagnose the physical conditions and processes. One of such methods, solar atmospheric seismology, makes use of observed and theoretically predicted properties of waves to infer plasma and magnetic field properties. A recent development in solar atmospheric seismology consists in the use of inversion and model comparison methods based on Bayesian analysis. In this paper, the philosophy and methodology of Bayesian analysis are first explained. Then, we provide an account of what has been achieved so far from the application of these techniques to solar atmospheric seismology and a prospect of possible future extensions.

Bayesian Hierarchical Modeling for Big Data Fusion in Soil Hydrology

NASA Astrophysics Data System (ADS)

Mohanty, B.; Kathuria, D.; Katzfuss, M.

2016-12-01

Soil moisture datasets from remote sensing (RS) platforms (such as SMOS and SMAP) and reanalysis products from land surface models are typically available on a coarse spatial granularity of several square km. Ground based sensors on the other hand provide observations on a finer spatial scale (meter scale or less) but are sparsely available. Soil moisture is affected by high variability due to complex interactions between geologic, topographic, vegetation and atmospheric variables. Hydrologic processes usually occur at a scale of 1 km or less and therefore spatially ubiquitous and temporally periodic soil moisture products at this scale are required to aid local decision makers in agriculture, weather prediction and reservoir operations. Past literature has largely focused on downscaling RS soil moisture for a small extent of a field or a watershed and hence the applicability of such products has been limited. The present study employs a spatial Bayesian Hierarchical Model (BHM) to derive soil moisture products at a spatial scale of 1 km for the state of Oklahoma by fusing point scale Mesonet data and coarse scale RS data for soil moisture and its auxiliary covariates such as precipitation, topography, soil texture and vegetation. It is seen that the BHM model handles change of support problems easily while performing accurate uncertainty quantification arising from measurement errors and imperfect retrieval algorithms. The computational challenge arising due to the large number of measurements is tackled by utilizing basis function approaches and likelihood approximations. The BHM model can be considered as a complex Bayesian extension of traditional geostatistical prediction methods (such as Kriging) for large datasets in the presence of uncertainties.

Filling gaps in notification data: a model-based approach applied to travel related campylobacteriosis cases in New Zealand.

PubMed

Amene, E; Horn, B; Pirie, R; Lake, R; Döpfer, D

2016-09-06

Data containing notified cases of disease are often compromised by incomplete or partial information related to individual cases. In an effort to enhance the value of information from enteric disease notifications in New Zealand, this study explored the use of Bayesian and Multiple Imputation (MI) models to fill risk factor data gaps. As a test case, overseas travel as a risk factor for infection with campylobacteriosis has been examined. Two methods, namely Bayesian Specification (BAS) and Multiple Imputation (MI), were compared regarding predictive performance for various levels of artificially induced missingness of overseas travel status in campylobacteriosis notification data. Predictive performance of the models was assessed through the Brier Score, the Area Under the ROC Curve and the Percent Bias of regression coefficients. Finally, the best model was selected and applied to predict missing overseas travel status of campylobacteriosis notifications. While no difference was observed in the predictive performance of the BAS and MI methods at a lower rate of missingness (<10 %), but the BAS approach performed better than MI at a higher rate of missingness (50 %, 65 %, 80 %). The estimated proportion (95 % Credibility Intervals) of travel related cases was greatest in highly urban District Health Boards (DHBs) in Counties Manukau, Auckland and Waitemata, at 0.37 (0.12, 0.57), 0.33 (0.13, 0.55) and 0.28 (0.10, 0.49), whereas the lowest proportion was estimated for more rural West Coast, Northland and Tairawhiti DHBs at 0.02 (0.01, 0.05), 0.03 (0.01, 0.08) and 0.04 (0.01, 0.06), respectively. The national rate of travel related campylobacteriosis cases was estimated at 0.16 (0.02, 0.48). The use of BAS offers a flexible approach to data augmentation particularly when the missing rate is very high and when the Missing At Random (MAR) assumption holds. High rates of travel associated cases in urban regions of New Zealand predicted by this approach are plausible given the high rate of travel in these regions, including destinations with higher risk of infection. The added advantage of using a Bayesian approach is that the model's prediction can be improved whenever new information becomes available.

Reducing the Uncertainty in Atlantic Meridional Overturning Circulation Projections Using Bayesian Model Averaging

NASA Astrophysics Data System (ADS)

Olson, R.; An, S. I.

2016-12-01

Atlantic Meridional Overturning Circulation (AMOC) in the ocean might slow down in the future, which can lead to a host of climatic effects in North Atlantic and throughout the world. Despite improvements in climate models and availability of new observations, AMOC projections remain uncertain. Here we constrain CMIP5 multi-model ensemble output with observations of a recently developed AMOC index to provide improved Bayesian predictions of future AMOC. Specifically, we first calculate yearly AMOC index loosely based on Rahmstorf et al. (2015) for years 1880—2004 for both observations, and the CMIP5 models for which relevant output is available. We then assign a weight to each model based on a Bayesian Model Averaging method that accounts for differential model skill in terms of both mean state and variability. We include the temporal autocorrelation in climate model errors, and account for the uncertainty in the parameters of our statistical model. We use the weights to provide future weighted projections of AMOC, and compare them to un-weighted ones. Our projections use bootstrapping to account for uncertainty in internal AMOC variability. We also perform spectral and other statistical analyses to show that AMOC index variability, both in models and in observations, is consistent with red noise. Our results improve on and complement previous work by using a new ensemble of climate models, a different observational metric, and an improved Bayesian weighting method that accounts for differential model skill at reproducing internal variability. Reference: Rahmstorf, S., Box, J. E., Feulner, G., Mann, M. E., Robinson, A., Rutherford, S., & Schaffernicht, E. J. (2015). Exceptional twentieth-century slowdown in atlantic ocean overturning circulation. Nature Climate Change, 5(5), 475-480. doi:10.1038/nclimate2554

Evidence cross-validation and Bayesian inference of MAST plasma equilibria

DOE Office of Scientific and Technical Information (OSTI.GOV)

Nessi, G. T. von; Hole, M. J.; Svensson, J.

2012-01-15

In this paper, current profiles for plasma discharges on the mega-ampere spherical tokamak are directly calculated from pickup coil, flux loop, and motional-Stark effect observations via methods based in the statistical theory of Bayesian analysis. By representing toroidal plasma current as a series of axisymmetric current beams with rectangular cross-section and inferring the current for each one of these beams, flux-surface geometry and q-profiles are subsequently calculated by elementary application of Biot-Savart's law. The use of this plasma model in the context of Bayesian analysis was pioneered by Svensson and Werner on the joint-European tokamak [Svensson and Werner,Plasma Phys. Controlledmore » Fusion 50(8), 085002 (2008)]. In this framework, linear forward models are used to generate diagnostic predictions, and the probability distribution for the currents in the collection of plasma beams was subsequently calculated directly via application of Bayes' formula. In this work, we introduce a new diagnostic technique to identify and remove outlier observations associated with diagnostics falling out of calibration or suffering from an unidentified malfunction. These modifications enable a good agreement between Bayesian inference of the last-closed flux-surface with other corroborating data, such as that from force balance considerations using EFIT++[Appel et al., ''A unified approach to equilibrium reconstruction'' Proceedings of the 33rd EPS Conference on Plasma Physics (Rome, Italy, 2006)]. In addition, this analysis also yields errors on the plasma current profile and flux-surface geometry as well as directly predicting the Shafranov shift of the plasma core.« less

Potential uncertainty reduction in model-averaged benchmark dose estimates informed by an additional dose study.

PubMed

Shao, Kan; Small, Mitchell J

2011-10-01

A methodology is presented for assessing the information value of an additional dosage experiment in existing bioassay studies. The analysis demonstrates the potential reduction in the uncertainty of toxicity metrics derived from expanded studies, providing insights for future studies. Bayesian methods are used to fit alternative dose-response models using Markov chain Monte Carlo (MCMC) simulation for parameter estimation and Bayesian model averaging (BMA) is used to compare and combine the alternative models. BMA predictions for benchmark dose (BMD) are developed, with uncertainty in these predictions used to derive the lower bound BMDL. The MCMC and BMA results provide a basis for a subsequent Monte Carlo analysis that backcasts the dosage where an additional test group would have been most beneficial in reducing the uncertainty in the BMD prediction, along with the magnitude of the expected uncertainty reduction. Uncertainty reductions are measured in terms of reduced interval widths of predicted BMD values and increases in BMDL values that occur as a result of this reduced uncertainty. The methodology is illustrated using two existing data sets for TCDD carcinogenicity, fitted with two alternative dose-response models (logistic and quantal-linear). The example shows that an additional dose at a relatively high value would have been most effective for reducing the uncertainty in BMA BMD estimates, with predicted reductions in the widths of uncertainty intervals of approximately 30%, and expected increases in BMDL values of 5-10%. The results demonstrate that dose selection for studies that subsequently inform dose-response models can benefit from consideration of how these models will be fit, combined, and interpreted. © 2011 Society for Risk Analysis.

ChemStable: a web server for rule-embedded naïve Bayesian learning approach to predict compound stability.

PubMed

Liu, Zhihong; Zheng, Minghao; Yan, Xin; Gu, Qiong; Gasteiger, Johann; Tijhuis, Johan; Maas, Peter; Li, Jiabo; Xu, Jun

2014-09-01

Predicting compound chemical stability is important because unstable compounds can lead to either false positive or to false negative conclusions in bioassays. Experimental data (COMDECOM) measured from DMSO/H2O solutions stored at 50 °C for 105 days were used to predicted stability by applying rule-embedded naïve Bayesian learning, based upon atom center fragment (ACF) features. To build the naïve Bayesian classifier, we derived ACF features from 9,746 compounds in the COMDECOM dataset. By recursively applying naïve Bayesian learning from the data set, each ACF is assigned with an expected stable probability (p(s)) and an unstable probability (p(uns)). 13,340 ACFs, together with their p(s) and p(uns) data, were stored in a knowledge base for use by the Bayesian classifier. For a given compound, its ACFs were derived from its structure connection table with the same protocol used to drive ACFs from the training data. Then, the Bayesian classifier assigned p(s) and p(uns) values to the compound ACFs by a structural pattern recognition algorithm, which was implemented in-house. Compound instability is calculated, with Bayes' theorem, based upon the p(s) and p(uns) values of the compound ACFs. We were able to achieve performance with an AUC value of 84% and a tenfold cross validation accuracy of 76.5%. To reduce false negatives, a rule-based approach has been embedded in the classifier. The rule-based module allows the program to improve its predictivity by expanding its compound instability knowledge base, thus further reducing the possibility of false negatives. To our knowledge, this is the first in silico prediction service for the prediction of the stabilities of organic compounds.

Dynamic Bayesian Network Modeling of the Interplay between EGFR and Hedgehog Signaling.

PubMed

Fröhlich, Holger; Bahamondez, Gloria; Götschel, Frank; Korf, Ulrike

2015-01-01

Aberrant activation of sonic Hegdehog (SHH) signaling has been found to disrupt cellular differentiation in many human cancers and to increase proliferation. The SHH pathway is known to cross-talk with EGFR dependent signaling. Recent studies experimentally addressed this interplay in Daoy cells, which are presumable a model system for medulloblastoma, a highly malignant brain tumor that predominately occurs in children. Currently ongoing are several clinical trials for different solid cancers, which are designed to validate the clinical benefits of targeting the SHH in combination with other pathways. This has motivated us to investigate interactions between EGFR and SHH dependent signaling in greater depth. To our knowledge, there is no mathematical model describing the interplay between EGFR and SHH dependent signaling in medulloblastoma so far. Here we come up with a fully probabilistic approach using Dynamic Bayesian Networks (DBNs). To build our model, we made use of literature based knowledge describing SHH and EGFR signaling and integrated gene expression (Illumina) and cellular location dependent time series protein expression data (Reverse Phase Protein Arrays). We validated our model by sub-sampling training data and making Bayesian predictions on the left out test data. Our predictions focusing on key transcription factors and p70S6K, showed a high level of concordance with experimental data. Furthermore, the stability of our model was tested by a parametric bootstrap approach. Stable network features were in agreement with published data. Altogether we believe that our model improved our understanding of the interplay between two highly oncogenic signaling pathways in Daoy cells. This may open new perspectives for the future therapy of Hedghog/EGF-dependent solid tumors.

Bayesian model evidence as a model evaluation metric

NASA Astrophysics Data System (ADS)

Guthke, Anneli; Höge, Marvin; Nowak, Wolfgang

2017-04-01

When building environmental systems models, we are typically confronted with the questions of how to choose an appropriate model (i.e., which processes to include or neglect) and how to measure its quality. Various metrics have been proposed that shall guide the modeller towards a most robust and realistic representation of the system under study. Criteria for evaluation often address aspects of accuracy (absence of bias) or of precision (absence of unnecessary variance) and need to be combined in a meaningful way in order to address the inherent bias-variance dilemma. We suggest using Bayesian model evidence (BME) as a model evaluation metric that implicitly performs a tradeoff between bias and variance. BME is typically associated with model weights in the context of Bayesian model averaging (BMA). However, it can also be seen as a model evaluation metric in a single-model context or in model comparison. It combines a measure for goodness of fit with a penalty for unjustifiable complexity. Unjustifiable refers to the fact that the appropriate level of model complexity is limited by the amount of information available for calibration. Derived in a Bayesian context, BME naturally accounts for measurement errors in the calibration data as well as for input and parameter uncertainty. BME is therefore perfectly suitable to assess model quality under uncertainty. We will explain in detail and with schematic illustrations what BME measures, i.e. how complexity is defined in the Bayesian setting and how this complexity is balanced with goodness of fit. We will further discuss how BME compares to other model evaluation metrics that address accuracy and precision such as the predictive logscore or other model selection criteria such as the AIC, BIC or KIC. Although computationally more expensive than other metrics or criteria, BME represents an appealing alternative because it provides a global measure of model quality. Even if not applicable to each and every case, we aim at stimulating discussion about how to judge the quality of hydrological models in the presence of uncertainty in general by dissecting the mechanism behind BME.

Bayesian calibration for electrochemical thermal model of lithium-ion cells

NASA Astrophysics Data System (ADS)

Tagade, Piyush; Hariharan, Krishnan S.; Basu, Suman; Verma, Mohan Kumar Singh; Kolake, Subramanya Mayya; Song, Taewon; Oh, Dukjin; Yeo, Taejung; Doo, Seokgwang

2016-07-01

Pseudo-two dimensional electrochemical thermal (P2D-ECT) model contains many parameters that are difficult to evaluate experimentally. Estimation of these model parameters is challenging due to computational cost and the transient model. Due to lack of complete physical understanding, this issue gets aggravated at extreme conditions like low temperature (LT) operations. This paper presents a Bayesian calibration framework for estimation of the P2D-ECT model parameters. The framework uses a matrix variate Gaussian process representation to obtain a computationally tractable formulation for calibration of the transient model. Performance of the framework is investigated for calibration of the P2D-ECT model across a range of temperatures (333 Ksbnd 263 K) and operating protocols. In the absence of complete physical understanding, the framework also quantifies structural uncertainty in the calibrated model. This information is used by the framework to test validity of the new physical phenomena before incorporation in the model. This capability is demonstrated by introducing temperature dependence on Bruggeman's coefficient and lithium plating formation at LT. With the incorporation of new physics, the calibrated P2D-ECT model accurately predicts the cell voltage with high confidence. The accurate predictions are used to obtain new insights into the low temperature lithium ion cell behavior.

A Bayesian Multilevel Model for Microcystin Prediction in Lakes of the Continental United States,

EPA Science Inventory

The frequency of cyanobacteria blooms in North American lakes is increasing. A major concernwith rising cyanobacteria blooms is microcystin, a common cyanobacterial hepatotoxin. Toexplore the conditions that promote high microcystin concentrations, we analyzed the US EPANational ...

Using Bayesian neural networks to classify forest scenes

NASA Astrophysics Data System (ADS)

Vehtari, Aki; Heikkonen, Jukka; Lampinen, Jouko; Juujarvi, Jouni

1998-10-01

We present results that compare the performance of Bayesian learning methods for neural networks on the task of classifying forest scenes into trees and background. Classification task is demanding due to the texture richness of the trees, occlusions of the forest scene objects and diverse lighting conditions under operation. This makes it difficult to determine which are optimal image features for the classification. A natural way to proceed is to extract many different types of potentially suitable features, and to evaluate their usefulness in later processing stages. One approach to cope with large number of features is to use Bayesian methods to control the model complexity. Bayesian learning uses a prior on model parameters, combines this with evidence from a training data, and the integrates over the resulting posterior to make predictions. With this method, we can use large networks and many features without fear of overfitting. For this classification task we compare two Bayesian learning methods for multi-layer perceptron (MLP) neural networks: (1) The evidence framework of MacKay uses a Gaussian approximation to the posterior weight distribution and maximizes with respect to hyperparameters. (2) In a Markov Chain Monte Carlo (MCMC) method due to Neal, the posterior distribution of the network parameters is numerically integrated using the MCMC method. As baseline classifiers for comparison we use (3) MLP early stop committee, (4) K-nearest-neighbor and (5) Classification And Regression Tree.

Advances in Parameter and Uncertainty Quantification Using Bayesian Hierarchical Techniques with a Spatially Referenced Watershed Model (Invited)

NASA Astrophysics Data System (ADS)

Alexander, R. B.; Boyer, E. W.; Schwarz, G. E.; Smith, R. A.

2013-12-01

Estimating water and material stores and fluxes in watershed studies is frequently complicated by uncertainties in quantifying hydrological and biogeochemical effects of factors such as land use, soils, and climate. Although these process-related effects are commonly measured and modeled in separate catchments, researchers are especially challenged by their complexity across catchments and diverse environmental settings, leading to a poor understanding of how model parameters and prediction uncertainties vary spatially. To address these concerns, we illustrate the use of Bayesian hierarchical modeling techniques with a dynamic version of the spatially referenced watershed model SPARROW (SPAtially Referenced Regression On Watershed attributes). The dynamic SPARROW model is designed to predict streamflow and other water cycle components (e.g., evapotranspiration, soil and groundwater storage) for monthly varying hydrological regimes, using mechanistic functions, mass conservation constraints, and statistically estimated parameters. In this application, the model domain includes nearly 30,000 NHD (National Hydrologic Data) stream reaches and their associated catchments in the Susquehanna River Basin. We report the results of our comparisons of alternative models of varying complexity, including models with different explanatory variables as well as hierarchical models that account for spatial and temporal variability in model parameters and variance (error) components. The model errors are evaluated for changes with season and catchment size and correlations in time and space. The hierarchical models consist of a two-tiered structure in which climate forcing parameters are modeled as random variables, conditioned on watershed properties. Quantification of spatial and temporal variations in the hydrological parameters and model uncertainties in this approach leads to more efficient (lower variance) and less biased model predictions throughout the river network. Moreover, predictions of water-balance components are reported according to probabilistic metrics (e.g., percentiles, prediction intervals) that include both parameter and model uncertainties. These improvements in predictions of streamflow dynamics can inform the development of more accurate predictions of spatial and temporal variations in biogeochemical stores and fluxes (e.g., nutrients and carbon) in watersheds.

Development and comparison of Bayesian modularization method in uncertainty assessment of hydrological models

NASA Astrophysics Data System (ADS)

Li, L.; Xu, C.-Y.; Engeland, K.

2012-04-01

With respect to model calibration, parameter estimation and analysis of uncertainty sources, different approaches have been used in hydrological models. Bayesian method is one of the most widely used methods for uncertainty assessment of hydrological models, which incorporates different sources of information into a single analysis through Bayesian theorem. However, none of these applications can well treat the uncertainty in extreme flows of hydrological models' simulations. This study proposes a Bayesian modularization method approach in uncertainty assessment of conceptual hydrological models by considering the extreme flows. It includes a comprehensive comparison and evaluation of uncertainty assessments by a new Bayesian modularization method approach and traditional Bayesian models using the Metropolis Hasting (MH) algorithm with the daily hydrological model WASMOD. Three likelihood functions are used in combination with traditional Bayesian: the AR (1) plus Normal and time period independent model (Model 1), the AR (1) plus Normal and time period dependent model (Model 2) and the AR (1) plus multi-normal model (Model 3). The results reveal that (1) the simulations derived from Bayesian modularization method are more accurate with the highest Nash-Sutcliffe efficiency value, and (2) the Bayesian modularization method performs best in uncertainty estimates of entire flows and in terms of the application and computational efficiency. The study thus introduces a new approach for reducing the extreme flow's effect on the discharge uncertainty assessment of hydrological models via Bayesian. Keywords: extreme flow, uncertainty assessment, Bayesian modularization, hydrological model, WASMOD

A New Method for Predicting Patient Survivorship Using Efficient Bayesian Network Learning

PubMed Central

Jiang, Xia; Xue, Diyang; Brufsky, Adam; Khan, Seema; Neapolitan, Richard

2014-01-01

The purpose of this investigation is to develop and evaluate a new Bayesian network (BN)-based patient survivorship prediction method. The central hypothesis is that the method predicts patient survivorship well, while having the capability to handle high-dimensional data and be incorporated into a clinical decision support system (CDSS). We have developed EBMC_Survivorship (EBMC_S), which predicts survivorship for each year individually. EBMC_S is based on the EBMC BN algorithm, which has been shown to handle high-dimensional data. BNs have excellent architecture for decision support systems. In this study, we evaluate EBMC_S using the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) dataset, which concerns breast tumors. A 5-fold cross-validation study indicates that EMBC_S performs better than the Cox proportional hazard model and is comparable to the random survival forest method. We show that EBMC_S provides additional information such as sensitivity analyses, which covariates predict each year, and yearly areas under the ROC curve (AUROCs). We conclude that our investigation supports the central hypothesis. PMID:24558297

A new method for predicting patient survivorship using efficient bayesian network learning.

PubMed

Jiang, Xia; Xue, Diyang; Brufsky, Adam; Khan, Seema; Neapolitan, Richard

2014-01-01

The purpose of this investigation is to develop and evaluate a new Bayesian network (BN)-based patient survivorship prediction method. The central hypothesis is that the method predicts patient survivorship well, while having the capability to handle high-dimensional data and be incorporated into a clinical decision support system (CDSS). We have developed EBMC_Survivorship (EBMC_S), which predicts survivorship for each year individually. EBMC_S is based on the EBMC BN algorithm, which has been shown to handle high-dimensional data. BNs have excellent architecture for decision support systems. In this study, we evaluate EBMC_S using the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) dataset, which concerns breast tumors. A 5-fold cross-validation study indicates that EMBC_S performs better than the Cox proportional hazard model and is comparable to the random survival forest method. We show that EBMC_S provides additional information such as sensitivity analyses, which covariates predict each year, and yearly areas under the ROC curve (AUROCs). We conclude that our investigation supports the central hypothesis.

Joint analysis of input and parametric uncertainties in watershed water quality modeling: A formal Bayesian approach

NASA Astrophysics Data System (ADS)

Han, Feng; Zheng, Yi

2018-06-01

Significant Input uncertainty is a major source of error in watershed water quality (WWQ) modeling. It remains challenging to address the input uncertainty in a rigorous Bayesian framework. This study develops the Bayesian Analysis of Input and Parametric Uncertainties (BAIPU), an approach for the joint analysis of input and parametric uncertainties through a tight coupling of Markov Chain Monte Carlo (MCMC) analysis and Bayesian Model Averaging (BMA). The formal likelihood function for this approach is derived considering a lag-1 autocorrelated, heteroscedastic, and Skew Exponential Power (SEP) distributed error model. A series of numerical experiments were performed based on a synthetic nitrate pollution case and on a real study case in the Newport Bay Watershed, California. The Soil and Water Assessment Tool (SWAT) and Differential Evolution Adaptive Metropolis (DREAM(ZS)) were used as the representative WWQ model and MCMC algorithm, respectively. The major findings include the following: (1) the BAIPU can be implemented and used to appropriately identify the uncertain parameters and characterize the predictive uncertainty; (2) the compensation effect between the input and parametric uncertainties can seriously mislead the modeling based management decisions, if the input uncertainty is not explicitly accounted for; (3) the BAIPU accounts for the interaction between the input and parametric uncertainties and therefore provides more accurate calibration and uncertainty results than a sequential analysis of the uncertainties; and (4) the BAIPU quantifies the credibility of different input assumptions on a statistical basis and can be implemented as an effective inverse modeling approach to the joint inference of parameters and inputs.

Predicting individual brain functional connectivity using a Bayesian hierarchical model.

PubMed

Dai, Tian; Guo, Ying

2017-02-15

Network-oriented analysis of functional magnetic resonance imaging (fMRI), especially resting-state fMRI, has revealed important association between abnormal connectivity and brain disorders such as schizophrenia, major depression and Alzheimer's disease. Imaging-based brain connectivity measures have become a useful tool for investigating the pathophysiology, progression and treatment response of psychiatric disorders and neurodegenerative diseases. Recent studies have started to explore the possibility of using functional neuroimaging to help predict disease progression and guide treatment selection for individual patients. These studies provide the impetus to develop statistical methodology that would help provide predictive information on disease progression-related or treatment-related changes in neural connectivity. To this end, we propose a prediction method based on Bayesian hierarchical model that uses individual's baseline fMRI scans, coupled with relevant subject characteristics, to predict the individual's future functional connectivity. A key advantage of the proposed method is that it can improve the accuracy of individualized prediction of connectivity by combining information from both group-level connectivity patterns that are common to subjects with similar characteristics as well as individual-level connectivity features that are particular to the specific subject. Furthermore, our method also offers statistical inference tools such as predictive intervals that help quantify the uncertainty or variability of the predicted outcomes. The proposed prediction method could be a useful approach to predict the changes in individual patient's brain connectivity with the progression of a disease. It can also be used to predict a patient's post-treatment brain connectivity after a specified treatment regimen. Another utility of the proposed method is that it can be applied to test-retest imaging data to develop a more reliable estimator for individual functional connectivity. We show there exists a nice connection between our proposed estimator and a recently developed shrinkage estimator of connectivity measures in the neuroimaging community. We develop an expectation-maximization (EM) algorithm for estimation of the proposed Bayesian hierarchical model. Simulations studies are performed to evaluate the accuracy of our proposed prediction methods. We illustrate the application of the methods with two data examples: the longitudinal resting-state fMRI from ADNI2 study and the test-retest fMRI data from Kirby21 study. In both the simulation studies and the fMRI data applications, we demonstrate that the proposed methods provide more accurate prediction and more reliable estimation of individual functional connectivity as compared with alternative methods. Copyright © 2017 Elsevier Inc. All rights reserved.

Spatial Attention, Motor Intention, and Bayesian Cue Predictability in the Human Brain.

PubMed

Kuhns, Anna B; Dombert, Pascasie L; Mengotti, Paola; Fink, Gereon R; Vossel, Simone

2017-05-24

Predictions about upcoming events influence how we perceive and respond to our environment. There is increasing evidence that predictions may be generated based upon previous observations following Bayesian principles, but little is known about the underlying cortical mechanisms and their specificity for different cognitive subsystems. The present study aimed at identifying common and distinct neural signatures of predictive processing in the spatial attentional and motor intentional system. Twenty-three female and male healthy human volunteers performed two probabilistic cueing tasks with either spatial or motor cues while lying in the fMRI scanner. In these tasks, the percentage of cue validity changed unpredictably over time. Trialwise estimates of cue predictability were derived from a Bayesian observer model of behavioral responses. These estimates were included as parametric regressors for analyzing the BOLD time series. Parametric effects of cue predictability in valid and invalid trials were considered to reflect belief updating by precision-weighted prediction errors. The brain areas exhibiting predictability-dependent effects dissociated between the spatial attention and motor intention task, with the right temporoparietal cortex being involved during spatial attention and the left angular gyrus and anterior cingulate cortex during motor intention. Connectivity analyses revealed that all three areas showed predictability-dependent coupling with the right hippocampus. These results suggest that precision-weighted prediction errors of stimulus locations and motor responses are encoded in distinct brain regions, but that crosstalk with the hippocampus may be necessary to integrate new trialwise outcomes in both cognitive systems. SIGNIFICANCE STATEMENT The brain is able to infer the environments' statistical structure and responds strongly to expectancy violations. In the spatial attentional domain, it has been shown that parts of the attentional networks are sensitive to the predictability of stimuli. It remains unknown, however, whether these effects are ubiquitous or if they are specific for different cognitive systems. The present study compared the influence of model-derived cue predictability on brain activity in the spatial attentional and motor intentional system. We identified areas with distinct predictability-dependent activation for spatial attention and motor intention, but also common connectivity changes of these regions with the hippocampus. These findings provide novel insights into the generality and specificity of predictive processing signatures in the human brain. Copyright © 2017 the authors 0270-6474/17/375334-11$15.00/0.

Robust Bayesian Experimental Design for Conceptual Model Discrimination

NASA Astrophysics Data System (ADS)

Pham, H. V.; Tsai, F. T. C.

2015-12-01

A robust Bayesian optimal experimental design under uncertainty is presented to provide firm information for model discrimination, given the least number of pumping wells and observation wells. Firm information is the maximum information of a system can be guaranteed from an experimental design. The design is based on the Box-Hill expected entropy decrease (EED) before and after the experiment design and the Bayesian model averaging (BMA) framework. A max-min programming is introduced to choose the robust design that maximizes the minimal Box-Hill EED subject to that the highest expected posterior model probability satisfies a desired probability threshold. The EED is calculated by the Gauss-Hermite quadrature. The BMA method is used to predict future observations and to quantify future observation uncertainty arising from conceptual and parametric uncertainties in calculating EED. Monte Carlo approach is adopted to quantify the uncertainty in the posterior model probabilities. The optimal experimental design is tested by a synthetic 5-layer anisotropic confined aquifer. Nine conceptual groundwater models are constructed due to uncertain geological architecture and boundary condition. High-performance computing is used to enumerate all possible design solutions in order to identify the most plausible groundwater model. Results highlight the impacts of scedasticity in future observation data as well as uncertainty sources on potential pumping and observation locations.

A Bayesian Ensemble Approach for Epidemiological Projections

PubMed Central

Lindström, Tom; Tildesley, Michael; Webb, Colleen

2015-01-01

Mathematical models are powerful tools for epidemiology and can be used to compare control actions. However, different models and model parameterizations may provide different prediction of outcomes. In other fields of research, ensemble modeling has been used to combine multiple projections. We explore the possibility of applying such methods to epidemiology by adapting Bayesian techniques developed for climate forecasting. We exemplify the implementation with single model ensembles based on different parameterizations of the Warwick model run for the 2001 United Kingdom foot and mouth disease outbreak and compare the efficacy of different control actions. This allows us to investigate the effect that discrepancy among projections based on different modeling assumptions has on the ensemble prediction. A sensitivity analysis showed that the choice of prior can have a pronounced effect on the posterior estimates of quantities of interest, in particular for ensembles with large discrepancy among projections. However, by using a hierarchical extension of the method we show that prior sensitivity can be circumvented. We further extend the method to include a priori beliefs about different modeling assumptions and demonstrate that the effect of this can have different consequences depending on the discrepancy among projections. We propose that the method is a promising analytical tool for ensemble modeling of disease outbreaks. PMID:25927892

Nonlinear finite element model updating for damage identification of civil structures using batch Bayesian estimation

NASA Astrophysics Data System (ADS)

Ebrahimian, Hamed; Astroza, Rodrigo; Conte, Joel P.; de Callafon, Raymond A.

2017-02-01

This paper presents a framework for structural health monitoring (SHM) and damage identification of civil structures. This framework integrates advanced mechanics-based nonlinear finite element (FE) modeling and analysis techniques with a batch Bayesian estimation approach to estimate time-invariant model parameters used in the FE model of the structure of interest. The framework uses input excitation and dynamic response of the structure and updates a nonlinear FE model of the structure to minimize the discrepancies between predicted and measured response time histories. The updated FE model can then be interrogated to detect, localize, classify, and quantify the state of damage and predict the remaining useful life of the structure. As opposed to recursive estimation methods, in the batch Bayesian estimation approach, the entire time history of the input excitation and output response of the structure are used as a batch of data to estimate the FE model parameters through a number of iterations. In the case of non-informative prior, the batch Bayesian method leads to an extended maximum likelihood (ML) estimation method to estimate jointly time-invariant model parameters and the measurement noise amplitude. The extended ML estimation problem is solved efficiently using a gradient-based interior-point optimization algorithm. Gradient-based optimization algorithms require the FE response sensitivities with respect to the model parameters to be identified. The FE response sensitivities are computed accurately and efficiently using the direct differentiation method (DDM). The estimation uncertainties are evaluated based on the Cramer-Rao lower bound (CRLB) theorem by computing the exact Fisher Information matrix using the FE response sensitivities with respect to the model parameters. The accuracy of the proposed uncertainty quantification approach is verified using a sampling approach based on the unscented transformation. Two validation studies, based on realistic structural FE models of a bridge pier and a moment resisting steel frame, are performed to validate the performance and accuracy of the presented nonlinear FE model updating approach and demonstrate its application to SHM. These validation studies show the excellent performance of the proposed framework for SHM and damage identification even in the presence of high measurement noise and/or way-out initial estimates of the model parameters. Furthermore, the detrimental effects of the input measurement noise on the performance of the proposed framework are illustrated and quantified through one of the validation studies.

[How exactly can we predict the prognosis of COPD].

PubMed

Atiş, Sibel; Kanik, Arzu; Ozgür, Eylem Sercan; Eker, Suzan; Tümkaya, Münir; Ozge, Cengiz

2009-01-01

Predictive models play a pivotal role in the provision of accurate and useful probabilistic assessments of clinical outcomes in chronic diseases. This study was aimed to develop a dedicated prognostic index for quantifying progression risk in chronic obstructive pulmonary disease (COPD). Data were collected prospectively from 75 COPD patients during a three years period. A predictive model of progression risk of COPD was developed using Bayesian logistic regression analysis by Markov chain Monte Carlo method. One-year cycles were used for the disease progression in this model. Primary end points for progression were impairment in basal dyspne index (BDI) score, FEV(1) decline, and exacerbation frequency in last three years. Time-varying covariates age, smoking, body mass index (BMI), severity of disease according to GOLD, PaO2, PaCO(2), IC, RV/TLC, DLCO were used under the study. The mean age was 57.1 + or - 8.1. BDI were strongly correlated with exacerbation frequency (p= 0.001) but not with FEV(1) decline. BMI was found to be a predictor factor for impairment in BDI (p= 0.03). The following independent risk factors were significant to predict exacerbation frequency: GOLD staging (OR for GOLD I vs. II and III = 2.3 and 4.0), hypoxemia (OR for mild vs moderate and severe = 2.1 and 5.1) and hyperinflation (OR= 1.6). PaO2 (p= 0.026), IC (p= 0.02) and RV/TLC (p= 0.03) were found to be predictive factors for FEV(1) decline. The model estimated BDI, lung function and exacerbation frequency at the last time point by testing initial data of three years with 95% reliability (p< 0.001). Accordingly, this model was evaluated as confident of 95% for assessing the future status of COPD patients. Using Bayesian predictive models, it was possible to develop a risk-stratification index that accurately predicted progression of COPD. This model can provide decision-making about future in COPD patients with high reliability looking clinical data of beginning.

Hierarchical Bayesian Markov switching models with application to predicting spawning success of shovelnose sturgeon

USGS Publications Warehouse

Holan, S.H.; Davis, G.M.; Wildhaber, M.L.; DeLonay, A.J.; Papoulias, D.M.

2009-01-01

The timing of spawning in fish is tightly linked to environmental factors; however, these factors are not very well understood for many species. Specifically, little information is available to guide recruitment efforts for endangered species such as the sturgeon. Therefore, we propose a Bayesian hierarchical model for predicting the success of spawning of the shovelnose sturgeon which uses both biological and behavioural (longitudinal) data. In particular, we use data that were produced from a tracking study that was conducted in the Lower Missouri River. The data that were produced from this study consist of biological variables associated with readiness to spawn along with longitudinal behavioural data collected by using telemetry and archival data storage tags. These high frequency data are complex both biologically and in the underlying behavioural process. To accommodate such complexity we developed a hierarchical linear regression model that uses an eigenvalue predictor, derived from the transition probability matrix of a two-state Markov switching model with generalized auto-regressive conditional heteroscedastic dynamics. Finally, to minimize the computational burden that is associated with estimation of this model, a parallel computing approach is proposed. ?? Journal compilation 2009 Royal Statistical Society.

Predictive uncertainty analysis of plume distribution for geological carbon sequestration using sparse-grid Bayesian method

NASA Astrophysics Data System (ADS)

Shi, X.; Zhang, G.

2013-12-01

Because of the extensive computational burden, parametric uncertainty analyses are rarely conducted for geological carbon sequestration (GCS) process based multi-phase models. The difficulty of predictive uncertainty analysis for the CO2 plume migration in realistic GCS models is not only due to the spatial distribution of the caprock and reservoir (i.e. heterogeneous model parameters), but also because the GCS optimization estimation problem has multiple local minima due to the complex nonlinear multi-phase (gas and aqueous), and multi-component (water, CO2, salt) transport equations. The geological model built by Doughty and Pruess (2004) for the Frio pilot site (Texas) was selected and assumed to represent the 'true' system, which was composed of seven different facies (geological units) distributed among 10 layers. We chose to calibrate the permeabilities of these facies. Pressure and gas saturation values from this true model were then extracted and used as observations for subsequent model calibration. Random noise was added to the observations to approximate realistic field conditions. Each simulation of the model lasts about 2 hours. In this study, we develop a new approach that improves computational efficiency of Bayesian inference by constructing a surrogate system based on an adaptive sparse-grid stochastic collocation method. This surrogate response surface global optimization algorithm is firstly used to calibrate the model parameters, then prediction uncertainty of the CO2 plume position is quantified due to the propagation from parametric uncertainty in the numerical experiments, which is also compared to the actual plume from the 'true' model. Results prove that the approach is computationally efficient for multi-modal optimization and prediction uncertainty quantification for computationally expensive simulation models. Both our inverse methodology and findings can be broadly applicable to GCS in heterogeneous storage formations.

Priors in perception: Top-down modulation, Bayesian perceptual learning rate, and prediction error minimization.

PubMed

Hohwy, Jakob

2017-01-01

I discuss top-down modulation of perception in terms of a variable Bayesian learning rate, revealing a wide range of prior hierarchical expectations that can modulate perception. I then switch to the prediction error minimization framework and seek to conceive cognitive penetration specifically as prediction error minimization deviations from a variable Bayesian learning rate. This approach retains cognitive penetration as a category somewhat distinct from other top-down effects, and carves a reasonable route between penetrability and impenetrability. It prevents rampant, relativistic cognitive penetration of perception and yet is consistent with the continuity of cognition and perception. Copyright © 2016 Elsevier Inc. All rights reserved.

Bayesian probabilistic approach for inverse source determination from limited and noisy chemical or biological sensor concentration measurements

NASA Astrophysics Data System (ADS)

Yee, Eugene

2007-04-01

Although a great deal of research effort has been focused on the forward prediction of the dispersion of contaminants (e.g., chemical and biological warfare agents) released into the turbulent atmosphere, much less work has been directed toward the inverse prediction of agent source location and strength from the measured concentration, even though the importance of this problem for a number of practical applications is obvious. In general, the inverse problem of source reconstruction is ill-posed and unsolvable without additional information. It is demonstrated that a Bayesian probabilistic inferential framework provides a natural and logically consistent method for source reconstruction from a limited number of noisy concentration data. In particular, the Bayesian approach permits one to incorporate prior knowledge about the source as well as additional information regarding both model and data errors. The latter enables a rigorous determination of the uncertainty in the inference of the source parameters (e.g., spatial location, emission rate, release time, etc.), hence extending the potential of the methodology as a tool for quantitative source reconstruction. A model (or, source-receptor relationship) that relates the source distribution to the concentration data measured by a number of sensors is formulated, and Bayesian probability theory is used to derive the posterior probability density function of the source parameters. A computationally efficient methodology for determination of the likelihood function for the problem, based on an adjoint representation of the source-receptor relationship, is described. Furthermore, we describe the application of efficient stochastic algorithms based on Markov chain Monte Carlo (MCMC) for sampling from the posterior distribution of the source parameters, the latter of which is required to undertake the Bayesian computation. The Bayesian inferential methodology for source reconstruction is validated against real dispersion data for two cases involving contaminant dispersion in highly disturbed flows over urban and complex environments where the idealizations of horizontal homogeneity and/or temporal stationarity in the flow cannot be applied to simplify the problem. Furthermore, the methodology is applied to the case of reconstruction of multiple sources.

Modelling Trial-by-Trial Changes in the Mismatch Negativity

PubMed Central

Lieder, Falk; Daunizeau, Jean; Garrido, Marta I.; Friston, Karl J.; Stephan, Klaas E.

2013-01-01

The mismatch negativity (MMN) is a differential brain response to violations of learned regularities. It has been used to demonstrate that the brain learns the statistical structure of its environment and predicts future sensory inputs. However, the algorithmic nature of these computations and the underlying neurobiological implementation remain controversial. This article introduces a mathematical framework with which competing ideas about the computational quantities indexed by MMN responses can be formalized and tested against single-trial EEG data. This framework was applied to five major theories of the MMN, comparing their ability to explain trial-by-trial changes in MMN amplitude. Three of these theories (predictive coding, model adjustment, and novelty detection) were formalized by linking the MMN to different manifestations of the same computational mechanism: approximate Bayesian inference according to the free-energy principle. We thereby propose a unifying view on three distinct theories of the MMN. The relative plausibility of each theory was assessed against empirical single-trial MMN amplitudes acquired from eight healthy volunteers in a roving oddball experiment. Models based on the free-energy principle provided more plausible explanations of trial-by-trial changes in MMN amplitude than models representing the two more traditional theories (change detection and adaptation). Our results suggest that the MMN reflects approximate Bayesian learning of sensory regularities, and that the MMN-generating process adjusts a probabilistic model of the environment according to prediction errors. PMID:23436989

Using Bayesian methods to predict climate impacts on groundwater availability and agricultural production in Punjab, India

NASA Astrophysics Data System (ADS)

Russo, T. A.; Devineni, N.; Lall, U.

2015-12-01

Lasting success of the Green Revolution in Punjab, India relies on continued availability of local water resources. Supplying primarily rice and wheat for the rest of India, Punjab supports crop irrigation with a canal system and groundwater, which is vastly over-exploited. The detailed data required to physically model future impacts on water supplies agricultural production is not readily available for this region, therefore we use Bayesian methods to estimate hydrologic properties and irrigation requirements for an under-constrained mass balance model. Using measured values of historical precipitation, total canal water delivery, crop yield, and water table elevation, we present a method using a Markov chain Monte Carlo (MCMC) algorithm to solve for a distribution of values for each unknown parameter in a conceptual mass balance model. Due to heterogeneity across the state, and the resolution of input data, we estimate model parameters at the district-scale using spatial pooling. The resulting model is used to predict the impact of precipitation change scenarios on groundwater availability under multiple cropping options. Predicted groundwater declines vary across the state, suggesting that crop selection and water management strategies should be determined at a local scale. This computational method can be applied in data-scarce regions across the world, where water resource management is required to resolve competition between food security and available resources in a changing climate.

Bayesian inference of galaxy formation from the K-band luminosity function of galaxies: tensions between theory and observation

NASA Astrophysics Data System (ADS)

Lu, Yu; Mo, H. J.; Katz, Neal; Weinberg, Martin D.

2012-04-01

We conduct Bayesian model inferences from the observed K-band luminosity function of galaxies in the local Universe, using the semi-analytic model (SAM) of galaxy formation introduced in Lu et al. The prior distributions for the 14 free parameters include a large range of possible models. We find that some of the free parameters, e.g. the characteristic scales for quenching star formation in both high-mass and low-mass haloes, are already tightly constrained by the single data set. The posterior distribution includes the model parameters adopted in other SAMs. By marginalizing over the posterior distribution, we make predictions that include the full inferential uncertainties for the colour-magnitude relation, the Tully-Fisher relation, the conditional stellar mass function of galaxies in haloes of different masses, the H I mass function, the redshift evolution of the stellar mass function of galaxies and the global star formation history. Using posterior predictive checking with the available observational results, we find that the model family (i) predicts a Tully-Fisher relation that is curved; (ii) significantly overpredicts the satellite fraction; (iii) vastly overpredicts the H I mass function; (iv) predicts high-z stellar mass functions that have too many low-mass galaxies and too few high-mass ones and (v) predicts a redshift evolution of the stellar mass density and the star formation history that are in moderate disagreement. These results suggest that some important processes are still missing in the current model family, and we discuss a number of possible solutions to solve the discrepancies, such as interactions between galaxies and dark matter haloes, tidal stripping, the bimodal accretion of gas, preheating and a redshift-dependent initial mass function.

A novel Bayesian approach to quantify clinical variables and to determine their spectroscopic counterparts in 1H NMR metabonomic data

PubMed Central

Vehtari, Aki; Mäkinen, Ville-Petteri; Soininen, Pasi; Ingman, Petri; Mäkelä, Sanna M; Savolainen, Markku J; Hannuksela, Minna L; Kaski, Kimmo; Ala-Korpela, Mika

2007-01-01

Background A key challenge in metabonomics is to uncover quantitative associations between multidimensional spectroscopic data and biochemical measures used for disease risk assessment and diagnostics. Here we focus on clinically relevant estimation of lipoprotein lipids by 1H NMR spectroscopy of serum. Results A Bayesian methodology, with a biochemical motivation, is presented for a real 1H NMR metabonomics data set of 75 serum samples. Lipoprotein lipid concentrations were independently obtained for these samples via ultracentrifugation and specific biochemical assays. The Bayesian models were constructed by Markov chain Monte Carlo (MCMC) and they showed remarkably good quantitative performance, the predictive R-values being 0.985 for the very low density lipoprotein triglycerides (VLDL-TG), 0.787 for the intermediate, 0.943 for the low, and 0.933 for the high density lipoprotein cholesterol (IDL-C, LDL-C and HDL-C, respectively). The modelling produced a kernel-based reformulation of the data, the parameters of which coincided with the well-known biochemical characteristics of the 1H NMR spectra; particularly for VLDL-TG and HDL-C the Bayesian methodology was able to clearly identify the most characteristic resonances within the heavily overlapping information in the spectra. For IDL-C and LDL-C the resulting model kernels were more complex than those for VLDL-TG and HDL-C, probably reflecting the severe overlap of the IDL and LDL resonances in the 1H NMR spectra. Conclusion The systematic use of Bayesian MCMC analysis is computationally demanding. Nevertheless, the combination of high-quality quantification and the biochemical rationale of the resulting models is expected to be useful in the field of metabonomics. PMID:17493257

Bayesian evidence and predictivity of the inflationary paradigm

DOE Office of Scientific and Technical Information (OSTI.GOV)

Gubitosi, Giulia; Lagos, Macarena; Magueijo, João

In this paper we consider the issue of paradigm evaluation by applying Bayes' theorem along the following nested hierarchy of progressively more complex structures: i) parameter estimation (within a model), ii) model selection and comparison (within a paradigm), iii) paradigm evaluation. In such a hierarchy the Bayesian evidence works both as the posterior's normalization at a given level and as the likelihood function at the next level up. Whilst raising no objections to the standard application of the procedure at the two lowest levels, we argue that it should receive a considerable modification when evaluating paradigms, when testability and fittingmore » data are equally important. By considering toy models we illustrate how models and paradigms that are difficult to falsify are always favoured by the Bayes factor. We argue that the evidence for a paradigm should not only be high for a given dataset, but exceptional with respect to what it would have been, had the data been different. With this motivation we propose a measure which we term predictivity , as well as a prior to be incorporated into the Bayesian framework, penalising unpredictivity as much as not fitting data. We apply this measure to inflation seen as a whole, and to a scenario where a specific inflationary model is hypothetically deemed as the only one viable as a result of information alien to cosmology (e.g. Solar System gravity experiments, or particle physics input). We conclude that cosmic inflation is currently hard to falsify, but that this could change were external/additional information to cosmology to select one of its many models. We also compare this state of affairs to bimetric varying speed of light cosmology.« less

A Bayesian Multilevel Model for Microcystin Prediction in Lakes of the Continental United States.

EPA Science Inventory

The frequency of cyanobacteria blooms in North American lakes is increasing. A major concern with rising cyanobacteria blooms is microcystin, a common cyanobacterial hepatotoxin. To explore the conditions that promote high microcystin concentrations, we analyzed the US EPA Nation...

A Bayesian Multilevel Model for Microcystin Prediction in Lakes of the Continental United States

EPA Science Inventory

The frequency of cyanobacteria blooms in North American lakes is increasing. A major concern with rising cyanobacteria blooms is microcystin, a common cyanobacterial hepatotoxin. To explore the conditions that promote high microcystin concentrations, we analyzed the US EPA Nation...