RNA folding kinetics using Monte Carlo and Gillespie algorithms.

PubMed

Clote, Peter; Bayegan, Amir H

2018-04-01

RNA secondary structure folding kinetics is known to be important for the biological function of certain processes, such as the hok/sok system in E. coli. Although linear algebra provides an exact computational solution of secondary structure folding kinetics with respect to the Turner energy model for tiny ([Formula: see text]20 nt) RNA sequences, the folding kinetics for larger sequences can only be approximated by binning structures into macrostates in a coarse-grained model, or by repeatedly simulating secondary structure folding with either the Monte Carlo algorithm or the Gillespie algorithm. Here we investigate the relation between the Monte Carlo algorithm and the Gillespie algorithm. We prove that asymptotically, the expected time for a K-step trajectory of the Monte Carlo algorithm is equal to [Formula: see text] times that of the Gillespie algorithm, where [Formula: see text] denotes the Boltzmann expected network degree. If the network is regular (i.e. every node has the same degree), then the mean first passage time (MFPT) computed by the Monte Carlo algorithm is equal to MFPT computed by the Gillespie algorithm multiplied by [Formula: see text]; however, this is not true for non-regular networks. In particular, RNA secondary structure folding kinetics, as computed by the Monte Carlo algorithm, is not equal to the folding kinetics, as computed by the Gillespie algorithm, although the mean first passage times are roughly correlated. Simulation software for RNA secondary structure folding according to the Monte Carlo and Gillespie algorithms is publicly available, as is our software to compute the expected degree of the network of secondary structures of a given RNA sequence-see http://bioinformatics.bc.edu/clote/RNAexpNumNbors .

A Hierarchical Multivariate Bayesian Approach to Ensemble Model output Statistics in Atmospheric Prediction

DTIC Science & Technology

2017-09-01

efficacy of statistical post-processing methods downstream of these dynamical model components with a hierarchical multivariate Bayesian approach to...Bayesian hierarchical modeling, Markov chain Monte Carlo methods , Metropolis algorithm, machine learning, atmospheric prediction 15. NUMBER OF PAGES...scale processes. However, this dissertation explores the efficacy of statistical post-processing methods downstream of these dynamical model components

Mastering the game of Go with deep neural networks and tree search

NASA Astrophysics Data System (ADS)

Silver, David; Huang, Aja; Maddison, Chris J.; Guez, Arthur; Sifre, Laurent; van den Driessche, George; Schrittwieser, Julian; Antonoglou, Ioannis; Panneershelvam, Veda; Lanctot, Marc; Dieleman, Sander; Grewe, Dominik; Nham, John; Kalchbrenner, Nal; Sutskever, Ilya; Lillicrap, Timothy; Leach, Madeleine; Kavukcuoglu, Koray; Graepel, Thore; Hassabis, Demis

2016-01-01

The game of Go has long been viewed as the most challenging of classic games for artificial intelligence owing to its enormous search space and the difficulty of evaluating board positions and moves. Here we introduce a new approach to computer Go that uses ‘value networks’ to evaluate board positions and ‘policy networks’ to select moves. These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play. Without any lookahead search, the neural networks play Go at the level of state-of-the-art Monte Carlo tree search programs that simulate thousands of random games of self-play. We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks. Using this search algorithm, our program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0. This is the first time that a computer program has defeated a human professional player in the full-sized game of Go, a feat previously thought to be at least a decade away.

Mastering the game of Go with deep neural networks and tree search.

PubMed

Silver, David; Huang, Aja; Maddison, Chris J; Guez, Arthur; Sifre, Laurent; van den Driessche, George; Schrittwieser, Julian; Antonoglou, Ioannis; Panneershelvam, Veda; Lanctot, Marc; Dieleman, Sander; Grewe, Dominik; Nham, John; Kalchbrenner, Nal; Sutskever, Ilya; Lillicrap, Timothy; Leach, Madeleine; Kavukcuoglu, Koray; Graepel, Thore; Hassabis, Demis

2016-01-28

The game of Go has long been viewed as the most challenging of classic games for artificial intelligence owing to its enormous search space and the difficulty of evaluating board positions and moves. Here we introduce a new approach to computer Go that uses 'value networks' to evaluate board positions and 'policy networks' to select moves. These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play. Without any lookahead search, the neural networks play Go at the level of state-of-the-art Monte Carlo tree search programs that simulate thousands of random games of self-play. We also introduce a new search algorithm that combines Monte Carlo simulation with value and policy networks. Using this search algorithm, our program AlphaGo achieved a 99.8% winning rate against other Go programs, and defeated the human European Go champion by 5 games to 0. This is the first time that a computer program has defeated a human professional player in the full-sized game of Go, a feat previously thought to be at least a decade away.

Cell-veto Monte Carlo algorithm for long-range systems.

PubMed

Kapfer, Sebastian C; Krauth, Werner

2016-09-01

We present a rigorous efficient event-chain Monte Carlo algorithm for long-range interacting particle systems. Using a cell-veto scheme within the factorized Metropolis algorithm, we compute each single-particle move with a fixed number of operations. For slowly decaying potentials such as Coulomb interactions, screening line charges allow us to take into account periodic boundary conditions. We discuss the performance of the cell-veto Monte Carlo algorithm for general inverse-power-law potentials, and illustrate how it provides a new outlook on one of the prominent bottlenecks in large-scale atomistic Monte Carlo simulations.

Quantum speedup of Monte Carlo methods.

PubMed

Montanaro, Ashley

2015-09-08

Monte Carlo methods use random sampling to estimate numerical quantities which are hard to compute deterministically. One important example is the use in statistical physics of rapidly mixing Markov chains to approximately compute partition functions. In this work, we describe a quantum algorithm which can accelerate Monte Carlo methods in a very general setting. The algorithm estimates the expected output value of an arbitrary randomized or quantum subroutine with bounded variance, achieving a near-quadratic speedup over the best possible classical algorithm. Combining the algorithm with the use of quantum walks gives a quantum speedup of the fastest known classical algorithms with rigorous performance bounds for computing partition functions, which use multiple-stage Markov chain Monte Carlo techniques. The quantum algorithm can also be used to estimate the total variation distance between probability distributions efficiently.

Quantum speedup of Monte Carlo methods

PubMed Central

Montanaro, Ashley

2015-01-01

Monte Carlo methods use random sampling to estimate numerical quantities which are hard to compute deterministically. One important example is the use in statistical physics of rapidly mixing Markov chains to approximately compute partition functions. In this work, we describe a quantum algorithm which can accelerate Monte Carlo methods in a very general setting. The algorithm estimates the expected output value of an arbitrary randomized or quantum subroutine with bounded variance, achieving a near-quadratic speedup over the best possible classical algorithm. Combining the algorithm with the use of quantum walks gives a quantum speedup of the fastest known classical algorithms with rigorous performance bounds for computing partition functions, which use multiple-stage Markov chain Monte Carlo techniques. The quantum algorithm can also be used to estimate the total variation distance between probability distributions efficiently. PMID:26528079

Genetic algorithm learning in a New Keynesian macroeconomic setup.

PubMed

Hommes, Cars; Makarewicz, Tomasz; Massaro, Domenico; Smits, Tom

2017-01-01

In order to understand heterogeneous behavior amongst agents, empirical data from Learning-to-Forecast (LtF) experiments can be used to construct learning models. This paper follows up on Assenza et al. (2013) by using a Genetic Algorithms (GA) model to replicate the results from their LtF experiment. In this GA model, individuals optimize an adaptive, a trend following and an anchor coefficient in a population of general prediction heuristics. We replicate experimental treatments in a New-Keynesian environment with increasing complexity and use Monte Carlo simulations to investigate how well the model explains the experimental data. We find that the evolutionary learning model is able to replicate the three different types of behavior, i.e. convergence to steady state, stable oscillations and dampened oscillations in the treatments using one GA model. Heterogeneous behavior can thus be explained by an adaptive, anchor and trend extrapolating component and the GA model can be used to explain heterogeneous behavior in LtF experiments with different types of complexity.

The Rational Hybrid Monte Carlo algorithm

NASA Astrophysics Data System (ADS)

Clark, Michael

2006-12-01

The past few years have seen considerable progress in algorithmic development for the generation of gauge fields including the effects of dynamical fermions. The Rational Hybrid Monte Carlo (RHMC) algorithm, where Hybrid Monte Carlo is performed using a rational approximation in place the usual inverse quark matrix kernel is one of these developments. This algorithm has been found to be extremely beneficial in many areas of lattice QCD (chiral fermions, finite temperature, Wilson fermions etc.). We review the algorithm and some of these benefits, and we compare against other recent algorithm developements. We conclude with an update of the Berlin wall plot comparing costs of all popular fermion formulations.

Analytic continuation of quantum Monte Carlo data by stochastic analytical inference.

PubMed

Fuchs, Sebastian; Pruschke, Thomas; Jarrell, Mark

2010-05-01

We present an algorithm for the analytic continuation of imaginary-time quantum Monte Carlo data which is strictly based on principles of Bayesian statistical inference. Within this framework we are able to obtain an explicit expression for the calculation of a weighted average over possible energy spectra, which can be evaluated by standard Monte Carlo simulations, yielding as by-product also the distribution function as function of the regularization parameter. Our algorithm thus avoids the usual ad hoc assumptions introduced in similar algorithms to fix the regularization parameter. We apply the algorithm to imaginary-time quantum Monte Carlo data and compare the resulting energy spectra with those from a standard maximum-entropy calculation.

SU-F-T-619: Dose Evaluation of Specific Patient Plans Based On Monte Carlo Algorithm for a CyberKnife Stereotactic Radiosurgery System

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

Piao, J; PLA 302 Hospital, Beijing; Xu, S

2016-06-15

Purpose: This study will use Monte Carlo to simulate the Cyberknife system, and intend to develop the third-party tool to evaluate the dose verification of specific patient plans in TPS. Methods: By simulating the treatment head using the BEAMnrc and DOSXYZnrc software, the comparison between the calculated and measured data will be done to determine the beam parameters. The dose distribution calculated in the Raytracing, Monte Carlo algorithms of TPS (Multiplan Ver4.0.2) and in-house Monte Carlo simulation method for 30 patient plans, which included 10 head, lung and liver cases in each, were analyzed. The γ analysis with the combinedmore » 3mm/3% criteria would be introduced to quantitatively evaluate the difference of the accuracy between three algorithms. Results: More than 90% of the global error points were less than 2% for the comparison of the PDD and OAR curves after determining the mean energy and FWHM.The relative ideal Monte Carlo beam model had been established. Based on the quantitative evaluation of dose accuracy for three algorithms, the results of γ analysis shows that the passing rates (84.88±9.67% for head,98.83±1.05% for liver,98.26±1.87% for lung) of PTV in 30 plans between Monte Carlo simulation and TPS Monte Carlo algorithms were good. And the passing rates (95.93±3.12%,99.84±0.33% in each) of PTV in head and liver plans between Monte Carlo simulation and TPS Ray-tracing algorithms were also good. But the difference of DVHs in lung plans between Monte Carlo simulation and Ray-tracing algorithms was obvious, and the passing rate (51.263±38.964%) of γ criteria was not good. It is feasible that Monte Carlo simulation was used for verifying the dose distribution of patient plans. Conclusion: Monte Carlo simulation algorithm developed in the CyberKnife system of this study can be used as a reference tool for the third-party tool, which plays an important role in dose verification of patient plans. This work was supported in part by the grant from Chinese Natural Science Foundation (Grant No. 11275105). Thanks for the support from Accuray Corp.« less

Sampling algorithms for validation of supervised learning models for Ising-like systems

NASA Astrophysics Data System (ADS)

Portman, Nataliya; Tamblyn, Isaac

2017-12-01

In this paper, we build and explore supervised learning models of ferromagnetic system behavior, using Monte-Carlo sampling of the spin configuration space generated by the 2D Ising model. Given the enormous size of the space of all possible Ising model realizations, the question arises as to how to choose a reasonable number of samples that will form physically meaningful and non-intersecting training and testing datasets. Here, we propose a sampling technique called ;ID-MH; that uses the Metropolis-Hastings algorithm creating Markov process across energy levels within the predefined configuration subspace. We show that application of this method retains phase transitions in both training and testing datasets and serves the purpose of validation of a machine learning algorithm. For larger lattice dimensions, ID-MH is not feasible as it requires knowledge of the complete configuration space. As such, we develop a new ;block-ID; sampling strategy: it decomposes the given structure into square blocks with lattice dimension N ≤ 5 and uses ID-MH sampling of candidate blocks. Further comparison of the performance of commonly used machine learning methods such as random forests, decision trees, k nearest neighbors and artificial neural networks shows that the PCA-based Decision Tree regressor is the most accurate predictor of magnetizations of the Ising model. For energies, however, the accuracy of prediction is not satisfactory, highlighting the need to consider more algorithmically complex methods (e.g., deep learning).

Microgrid energy dispatching for industrial zones with renewable generations and electric vehicles via stochastic optimization and learning

NASA Astrophysics Data System (ADS)

Zhang, Kai; Li, Jingzhi; He, Zhubin; Yan, Wanfeng

2018-07-01

In this paper, a stochastic optimization framework is proposed to address the microgrid energy dispatching problem with random renewable generation and vehicle activity pattern, which is closer to the practical applications. The patterns of energy generation, consumption and storage availability are all random and unknown at the beginning, and the microgrid controller design (MCD) is formulated as a Markov decision process (MDP). Hence, an online learning-based control algorithm is proposed for the microgrid, which could adapt the control policy with increasing knowledge of the system dynamics and converges to the optimal algorithm. We adopt the linear approximation idea to decompose the original value functions as the summation of each per-battery value function. As a consequence, the computational complexity is significantly reduced from exponential growth to linear growth with respect to the size of battery states. Monte Carlo simulation of different scenarios demonstrates the effectiveness and efficiency of our algorithm.

Molecular Monte Carlo Simulations Using Graphics Processing Units: To Waste Recycle or Not?

PubMed

Kim, Jihan; Rodgers, Jocelyn M; Athènes, Manuel; Smit, Berend

2011-10-11

In the waste recycling Monte Carlo (WRMC) algorithm, (1) multiple trial states may be simultaneously generated and utilized during Monte Carlo moves to improve the statistical accuracy of the simulations, suggesting that such an algorithm may be well posed for implementation in parallel on graphics processing units (GPUs). In this paper, we implement two waste recycling Monte Carlo algorithms in CUDA (Compute Unified Device Architecture) using uniformly distributed random trial states and trial states based on displacement random-walk steps, and we test the methods on a methane-zeolite MFI framework system to evaluate their utility. We discuss the specific implementation details of the waste recycling GPU algorithm and compare the methods to other parallel algorithms optimized for the framework system. We analyze the relationship between the statistical accuracy of our simulations and the CUDA block size to determine the efficient allocation of the GPU hardware resources. We make comparisons between the GPU and the serial CPU Monte Carlo implementations to assess speedup over conventional microprocessors. Finally, we apply our optimized GPU algorithms to the important problem of determining free energy landscapes, in this case for molecular motion through the zeolite LTA.

Off-diagonal expansion quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Albash, Tameem; Wagenbreth, Gene; Hen, Itay

2017-12-01

We propose a Monte Carlo algorithm designed to simulate quantum as well as classical systems at equilibrium, bridging the algorithmic gap between quantum and classical thermal simulation algorithms. The method is based on a decomposition of the quantum partition function that can be viewed as a series expansion about its classical part. We argue that the algorithm not only provides a theoretical advancement in the field of quantum Monte Carlo simulations, but is optimally suited to tackle quantum many-body systems that exhibit a range of behaviors from "fully quantum" to "fully classical," in contrast to many existing methods. We demonstrate the advantages, sometimes by orders of magnitude, of the technique by comparing it against existing state-of-the-art schemes such as path integral quantum Monte Carlo and stochastic series expansion. We also illustrate how our method allows for the unification of quantum and classical thermal parallel tempering techniques into a single algorithm and discuss its practical significance.

Off-diagonal expansion quantum Monte Carlo.

PubMed

Albash, Tameem; Wagenbreth, Gene; Hen, Itay

2017-12-01

We propose a Monte Carlo algorithm designed to simulate quantum as well as classical systems at equilibrium, bridging the algorithmic gap between quantum and classical thermal simulation algorithms. The method is based on a decomposition of the quantum partition function that can be viewed as a series expansion about its classical part. We argue that the algorithm not only provides a theoretical advancement in the field of quantum Monte Carlo simulations, but is optimally suited to tackle quantum many-body systems that exhibit a range of behaviors from "fully quantum" to "fully classical," in contrast to many existing methods. We demonstrate the advantages, sometimes by orders of magnitude, of the technique by comparing it against existing state-of-the-art schemes such as path integral quantum Monte Carlo and stochastic series expansion. We also illustrate how our method allows for the unification of quantum and classical thermal parallel tempering techniques into a single algorithm and discuss its practical significance.

Multiscale Monte Carlo equilibration: Pure Yang-Mills theory

DOE PAGES

Endres, Michael G.; Brower, Richard C.; Orginos, Kostas; ...

2015-12-29

In this study, we present a multiscale thermalization algorithm for lattice gauge theory, which enables efficient parallel generation of uncorrelated gauge field configurations. The algorithm combines standard Monte Carlo techniques with ideas drawn from real space renormalization group and multigrid methods. We demonstrate the viability of the algorithm for pure Yang-Mills gauge theory for both heat bath and hybrid Monte Carlo evolution, and show that it ameliorates the problem of topological freezing up to controllable lattice spacing artifacts.

Collision of Physics and Software in the Monte Carlo Application Toolkit (MCATK)

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

Sweezy, Jeremy Ed

2016-01-21

The topic is presented in a series of slides organized as follows: MCATK overview, development strategy, available algorithms, problem modeling (sources, geometry, data, tallies), parallelism, miscellaneous tools/features, example MCATK application, recent areas of research, and summary and future work. MCATK is a C++ component-based Monte Carlo neutron-gamma transport software library with continuous energy neutron and photon transport. Designed to build specialized applications and to provide new functionality in existing general-purpose Monte Carlo codes like MCNP, it reads ACE formatted nuclear data generated by NJOY. The motivation behind MCATK was to reduce costs. MCATK physics involves continuous energy neutron & gammamore » transport with multi-temperature treatment, static eigenvalue (k eff and α) algorithms, time-dependent algorithm, and fission chain algorithms. MCATK geometry includes mesh geometries and solid body geometries. MCATK provides verified, unit-test Monte Carlo components, flexibility in Monte Carlo application development, and numerous tools such as geometry and cross section plotters.« less

pyNSMC: A Python Module for Null-Space Monte Carlo Uncertainty Analysis

NASA Astrophysics Data System (ADS)

White, J.; Brakefield, L. K.

2015-12-01

The null-space monte carlo technique is a non-linear uncertainty analyses technique that is well-suited to high-dimensional inverse problems. While the technique is powerful, the existing workflow for completing null-space monte carlo is cumbersome, requiring the use of multiple commandline utilities, several sets of intermediate files and even a text editor. pyNSMC is an open-source python module that automates the workflow of null-space monte carlo uncertainty analyses. The module is fully compatible with the PEST and PEST++ software suites and leverages existing functionality of pyEMU, a python framework for linear-based uncertainty analyses. pyNSMC greatly simplifies the existing workflow for null-space monte carlo by taking advantage of object oriented design facilities in python. The core of pyNSMC is the ensemble class, which draws and stores realized random vectors and also provides functionality for exporting and visualizing results. By relieving users of the tedium associated with file handling and command line utility execution, pyNSMC instead focuses the user on the important steps and assumptions of null-space monte carlo analysis. Furthermore, pyNSMC facilitates learning through flow charts and results visualization, which are available at many points in the algorithm. The ease-of-use of the pyNSMC workflow is compared to the existing workflow for null-space monte carlo for a synthetic groundwater model with hundreds of estimable parameters.

Evaluation of an analytic linear Boltzmann transport equation solver for high-density inhomogeneities

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

Lloyd, S. A. M.; Ansbacher, W.; Department of Physics and Astronomy, University of Victoria, Victoria, British Columbia V8W 3P6

2013-01-15

Purpose: Acuros external beam (Acuros XB) is a novel dose calculation algorithm implemented through the ECLIPSE treatment planning system. The algorithm finds a deterministic solution to the linear Boltzmann transport equation, the same equation commonly solved stochastically by Monte Carlo methods. This work is an evaluation of Acuros XB, by comparison with Monte Carlo, for dose calculation applications involving high-density materials. Existing non-Monte Carlo clinical dose calculation algorithms, such as the analytic anisotropic algorithm (AAA), do not accurately model dose perturbations due to increased electron scatter within high-density volumes. Methods: Acuros XB, AAA, and EGSnrc based Monte Carlo are usedmore » to calculate dose distributions from 18 MV and 6 MV photon beams delivered to a cubic water phantom containing a rectangular high density (4.0-8.0 g/cm{sup 3}) volume at its center. The algorithms are also used to recalculate a clinical prostate treatment plan involving a unilateral hip prosthesis, originally evaluated using AAA. These results are compared graphically and numerically using gamma-index analysis. Radio-chromic film measurements are presented to augment Monte Carlo and Acuros XB dose perturbation data. Results: Using a 2% and 1 mm gamma-analysis, between 91.3% and 96.8% of Acuros XB dose voxels containing greater than 50% the normalized dose were in agreement with Monte Carlo data for virtual phantoms involving 18 MV and 6 MV photons, stainless steel and titanium alloy implants and for on-axis and oblique field delivery. A similar gamma-analysis of AAA against Monte Carlo data showed between 80.8% and 87.3% agreement. Comparing Acuros XB and AAA evaluations of a clinical prostate patient plan involving a unilateral hip prosthesis, Acuros XB showed good overall agreement with Monte Carlo while AAA underestimated dose on the upstream medial surface of the prosthesis due to electron scatter from the high-density material. Film measurements support the dose perturbations demonstrated by Monte Carlo and Acuros XB data. Conclusions: Acuros XB is shown to perform as well as Monte Carlo methods and better than existing clinical algorithms for dose calculations involving high-density volumes.« less

SU-E-T-188: Film Dosimetry Verification of Monte Carlo Generated Electron Treatment Plans

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

Enright, S; Asprinio, A; Lu, L

2014-06-01

Purpose: The purpose of this study was to compare dose distributions from film measurements to Monte Carlo generated electron treatment plans. Irradiation with electrons offers the advantages of dose uniformity in the target volume and of minimizing the dose to deeper healthy tissue. Using the Monte Carlo algorithm will improve dose accuracy in regions with heterogeneities and irregular surfaces. Methods: Dose distributions from GafChromic{sup ™} EBT3 films were compared to dose distributions from the Electron Monte Carlo algorithm in the Eclipse{sup ™} radiotherapy treatment planning system. These measurements were obtained for 6MeV, 9MeV and 12MeV electrons at two depths. Allmore » phantoms studied were imported into Eclipse by CT scan. A 1 cm thick solid water template with holes for bonelike and lung-like plugs was used. Different configurations were used with the different plugs inserted into the holes. Configurations with solid-water plugs stacked on top of one another were also used to create an irregular surface. Results: The dose distributions measured from the film agreed with those from the Electron Monte Carlo treatment plan. Accuracy of Electron Monte Carlo algorithm was also compared to that of Pencil Beam. Dose distributions from Monte Carlo had much higher pass rates than distributions from Pencil Beam when compared to the film. The pass rate for Monte Carlo was in the 80%–99% range, where the pass rate for Pencil Beam was as low as 10.76%. Conclusion: The dose distribution from Monte Carlo agreed with the measured dose from the film. When compared to the Pencil Beam algorithm, pass rates for Monte Carlo were much higher. Monte Carlo should be used over Pencil Beam for regions with heterogeneities and irregular surfaces.« less

Renyi entanglement entropy of interacting fermions calculated using the continuous-time quantum Monte Carlo method.

PubMed

Wang, Lei; Troyer, Matthias

2014-09-12

We present a new algorithm for calculating the Renyi entanglement entropy of interacting fermions using the continuous-time quantum Monte Carlo method. The algorithm only samples the interaction correction of the entanglement entropy, which by design ensures the efficient calculation of weakly interacting systems. Combined with Monte Carlo reweighting, the algorithm also performs well for systems with strong interactions. We demonstrate the potential of this method by studying the quantum entanglement signatures of the charge-density-wave transition of interacting fermions on a square lattice.

Spectral Learning for Supervised Topic Models.

PubMed

Ren, Yong; Wang, Yining; Zhu, Jun

2018-03-01

Supervised topic models simultaneously model the latent topic structure of large collections of documents and a response variable associated with each document. Existing inference methods are based on variational approximation or Monte Carlo sampling, which often suffers from the local minimum defect. Spectral methods have been applied to learn unsupervised topic models, such as latent Dirichlet allocation (LDA), with provable guarantees. This paper investigates the possibility of applying spectral methods to recover the parameters of supervised LDA (sLDA). We first present a two-stage spectral method, which recovers the parameters of LDA followed by a power update method to recover the regression model parameters. Then, we further present a single-phase spectral algorithm to jointly recover the topic distribution matrix as well as the regression weights. Our spectral algorithms are provably correct and computationally efficient. We prove a sample complexity bound for each algorithm and subsequently derive a sufficient condition for the identifiability of sLDA. Thorough experiments on synthetic and real-world datasets verify the theory and demonstrate the practical effectiveness of the spectral algorithms. In fact, our results on a large-scale review rating dataset demonstrate that our single-phase spectral algorithm alone gets comparable or even better performance than state-of-the-art methods, while previous work on spectral methods has rarely reported such promising performance.

Event-chain Monte Carlo algorithms for three- and many-particle interactions

NASA Astrophysics Data System (ADS)

Harland, J.; Michel, M.; Kampmann, T. A.; Kierfeld, J.

2017-02-01

We generalize the rejection-free event-chain Monte Carlo algorithm from many-particle systems with pairwise interactions to systems with arbitrary three- or many-particle interactions. We introduce generalized lifting probabilities between particles and obtain a general set of equations for lifting probabilities, the solution of which guarantees maximal global balance. We validate the resulting three-particle event-chain Monte Carlo algorithms on three different systems by comparison with conventional local Monte Carlo simulations: i) a test system of three particles with a three-particle interaction that depends on the enclosed triangle area; ii) a hard-needle system in two dimensions, where needle interactions constitute three-particle interactions of the needle end points; iii) a semiflexible polymer chain with a bending energy, which constitutes a three-particle interaction of neighboring chain beads. The examples demonstrate that the generalization to many-particle interactions broadens the applicability of event-chain algorithms considerably.

Learning Kinetic Monte Carlo Models of Condensed Phase High Temperature Chemistry from Molecular Dynamics

NASA Astrophysics Data System (ADS)

Yang, Qian; Sing-Long, Carlos; Chen, Enze; Reed, Evan

2017-06-01

Complex chemical processes, such as the decomposition of energetic materials and the chemistry of planetary interiors, are typically studied using large-scale molecular dynamics simulations that run for weeks on high performance parallel machines. These computations may involve thousands of atoms forming hundreds of molecular species and undergoing thousands of reactions. It is natural to wonder whether this wealth of data can be utilized to build more efficient, interpretable, and predictive models. In this talk, we will use techniques from statistical learning to develop a framework for constructing Kinetic Monte Carlo (KMC) models from molecular dynamics data. We will show that our KMC models can not only extrapolate the behavior of the chemical system by as much as an order of magnitude in time, but can also be used to study the dynamics of entirely different chemical trajectories with a high degree of fidelity. Then, we will discuss three different methods for reducing our learned KMC models, including a new and efficient data-driven algorithm using L1-regularization. We demonstrate our framework throughout on a system of high-temperature high-pressure liquid methane, thought to be a major component of gas giant planetary interiors.

Scalable Domain Decomposed Monte Carlo Particle Transport

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

O'Brien, Matthew Joseph

2013-12-05

In this dissertation, we present the parallel algorithms necessary to run domain decomposed Monte Carlo particle transport on large numbers of processors (millions of processors). Previous algorithms were not scalable, and the parallel overhead became more computationally costly than the numerical simulation.

Implementation of Monte Carlo Dose calculation for CyberKnife treatment planning

NASA Astrophysics Data System (ADS)

Ma, C.-M.; Li, J. S.; Deng, J.; Fan, J.

2008-02-01

Accurate dose calculation is essential to advanced stereotactic radiosurgery (SRS) and stereotactic radiotherapy (SRT) especially for treatment planning involving heterogeneous patient anatomy. This paper describes the implementation of a fast Monte Carlo dose calculation algorithm in SRS/SRT treatment planning for the CyberKnife® SRS/SRT system. A superposition Monte Carlo algorithm is developed for this application. Photon mean free paths and interaction types for different materials and energies as well as the tracks of secondary electrons are pre-simulated using the MCSIM system. Photon interaction forcing and splitting are applied to the source photons in the patient calculation and the pre-simulated electron tracks are repeated with proper corrections based on the tissue density and electron stopping powers. Electron energy is deposited along the tracks and accumulated in the simulation geometry. Scattered and bremsstrahlung photons are transported, after applying the Russian roulette technique, in the same way as the primary photons. Dose calculations are compared with full Monte Carlo simulations performed using EGS4/MCSIM and the CyberKnife treatment planning system (TPS) for lung, head & neck and liver treatments. Comparisons with full Monte Carlo simulations show excellent agreement (within 0.5%). More than 10% differences in the target dose are found between Monte Carlo simulations and the CyberKnife TPS for SRS/SRT lung treatment while negligible differences are shown in head and neck and liver for the cases investigated. The calculation time using our superposition Monte Carlo algorithm is reduced up to 62 times (46 times on average for 10 typical clinical cases) compared to full Monte Carlo simulations. SRS/SRT dose distributions calculated by simple dose algorithms may be significantly overestimated for small lung target volumes, which can be improved by accurate Monte Carlo dose calculations.

Hybrid dose calculation: a dose calculation algorithm for microbeam radiation therapy

NASA Astrophysics Data System (ADS)

Donzelli, Mattia; Bräuer-Krisch, Elke; Oelfke, Uwe; Wilkens, Jan J.; Bartzsch, Stefan

2018-02-01

Microbeam radiation therapy (MRT) is still a preclinical approach in radiation oncology that uses planar micrometre wide beamlets with extremely high peak doses, separated by a few hundred micrometre wide low dose regions. Abundant preclinical evidence demonstrates that MRT spares normal tissue more effectively than conventional radiation therapy, at equivalent tumour control. In order to launch first clinical trials, accurate and efficient dose calculation methods are an inevitable prerequisite. In this work a hybrid dose calculation approach is presented that is based on a combination of Monte Carlo and kernel based dose calculation. In various examples the performance of the algorithm is compared to purely Monte Carlo and purely kernel based dose calculations. The accuracy of the developed algorithm is comparable to conventional pure Monte Carlo calculations. In particular for inhomogeneous materials the hybrid dose calculation algorithm out-performs purely convolution based dose calculation approaches. It is demonstrated that the hybrid algorithm can efficiently calculate even complicated pencil beam and cross firing beam geometries. The required calculation times are substantially lower than for pure Monte Carlo calculations.

Event-chain algorithm for the Heisenberg model: Evidence for z≃1 dynamic scaling.

PubMed

Nishikawa, Yoshihiko; Michel, Manon; Krauth, Werner; Hukushima, Koji

2015-12-01

We apply the event-chain Monte Carlo algorithm to the three-dimensional ferromagnetic Heisenberg model. The algorithm is rejection-free and also realizes an irreversible Markov chain that satisfies global balance. The autocorrelation functions of the magnetic susceptibility and the energy indicate a dynamical critical exponent z≈1 at the critical temperature, while that of the magnetization does not measure the performance of the algorithm. We show that the event-chain Monte Carlo algorithm substantially reduces the dynamical critical exponent from the conventional value of z≃2.

Data decomposition of Monte Carlo particle transport simulations via tally servers

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

Romano, Paul K.; Siegel, Andrew R.; Forget, Benoit

An algorithm for decomposing large tally data in Monte Carlo particle transport simulations is developed, analyzed, and implemented in a continuous-energy Monte Carlo code, OpenMC. The algorithm is based on a non-overlapping decomposition of compute nodes into tracking processors and tally servers. The former are used to simulate the movement of particles through the domain while the latter continuously receive and update tally data. A performance model for this approach is developed, suggesting that, for a range of parameters relevant to LWR analysis, the tally server algorithm should perform with minimal overhead on contemporary supercomputers. An implementation of the algorithmmore » in OpenMC is then tested on the Intrepid and Titan supercomputers, supporting the key predictions of the model over a wide range of parameters. We thus conclude that the tally server algorithm is a successful approach to circumventing classical on-node memory constraints en route to unprecedentedly detailed Monte Carlo reactor simulations.« less

A novel algorithm for solving the true coincident counting issues in Monte Carlo simulations for radiation spectroscopy.

PubMed

Guan, Fada; Johns, Jesse M; Vasudevan, Latha; Zhang, Guoqing; Tang, Xiaobin; Poston, John W; Braby, Leslie A

2015-06-01

Coincident counts can be observed in experimental radiation spectroscopy. Accurate quantification of the radiation source requires the detection efficiency of the spectrometer, which is often experimentally determined. However, Monte Carlo analysis can be used to supplement experimental approaches to determine the detection efficiency a priori. The traditional Monte Carlo method overestimates the detection efficiency as a result of omitting coincident counts caused mainly by multiple cascade source particles. In this study, a novel "multi-primary coincident counting" algorithm was developed using the Geant4 Monte Carlo simulation toolkit. A high-purity Germanium detector for ⁶⁰Co gamma-ray spectroscopy problems was accurately modeled to validate the developed algorithm. The simulated pulse height spectrum agreed well qualitatively with the measured spectrum obtained using the high-purity Germanium detector. The developed algorithm can be extended to other applications, with a particular emphasis on challenging radiation fields, such as counting multiple types of coincident radiations released from nuclear fission or used nuclear fuel.

Mosaicing of airborne LiDAR bathymetry strips based on Monte Carlo matching

NASA Astrophysics Data System (ADS)

Yang, Fanlin; Su, Dianpeng; Zhang, Kai; Ma, Yue; Wang, Mingwei; Yang, Anxiu

2017-09-01

This study proposes a new methodology for mosaicing airborne light detection and ranging (LiDAR) bathymetry (ALB) data based on Monte Carlo matching. Various errors occur in ALB data due to imperfect system integration and other interference factors. To account for these errors, a Monte Carlo matching algorithm based on a nonlinear least-squares adjustment model is proposed. First, the raw data of strip overlap areas were filtered according to their relative drift of depths. Second, a Monte Carlo model and nonlinear least-squares adjustment model were combined to obtain seven transformation parameters. Then, the multibeam bathymetric data were used to correct the initial strip during strip mosaicing. Finally, to evaluate the proposed method, the experimental results were compared with the results of the Iterative Closest Points (ICP) and three-dimensional Normal Distributions Transform (3D-NDT) algorithms. The results demonstrate that the algorithm proposed in this study is more robust and effective. When the quality of the raw data is poor, the Monte Carlo matching algorithm can still achieve centimeter-level accuracy for overlapping areas, which meets the accuracy of bathymetry required by IHO Standards for Hydrographic Surveys Special Publication No.44.

Scalable Domain Decomposed Monte Carlo Particle Transport

NASA Astrophysics Data System (ADS)

O'Brien, Matthew Joseph

In this dissertation, we present the parallel algorithms necessary to run domain decomposed Monte Carlo particle transport on large numbers of processors (millions of processors). Previous algorithms were not scalable, and the parallel overhead became more computationally costly than the numerical simulation. The main algorithms we consider are: • Domain decomposition of constructive solid geometry: enables extremely large calculations in which the background geometry is too large to fit in the memory of a single computational node. • Load Balancing: keeps the workload per processor as even as possible so the calculation runs efficiently. • Global Particle Find: if particles are on the wrong processor, globally resolve their locations to the correct processor based on particle coordinate and background domain. • Visualizing constructive solid geometry, sourcing particles, deciding that particle streaming communication is completed and spatial redecomposition. These algorithms are some of the most important parallel algorithms required for domain decomposed Monte Carlo particle transport. We demonstrate that our previous algorithms were not scalable, prove that our new algorithms are scalable, and run some of the algorithms up to 2 million MPI processes on the Sequoia supercomputer.

exprso: an R-package for the rapid implementation of machine learning algorithms.

PubMed

Quinn, Thomas; Tylee, Daniel; Glatt, Stephen

2016-01-01

Machine learning plays a major role in many scientific investigations. However, non-expert programmers may struggle to implement the elaborate pipelines necessary to build highly accurate and generalizable models. We introduce exprso , a new R package that is an intuitive machine learning suite designed specifically for non-expert programmers. Built initially for the classification of high-dimensional data, exprso uses an object-oriented framework to encapsulate a number of common analytical methods into a series of interchangeable modules. This includes modules for feature selection, classification, high-throughput parameter grid-searching, elaborate cross-validation schemes (e.g., Monte Carlo and nested cross-validation), ensemble classification, and prediction. In addition, exprso also supports multi-class classification (through the 1-vs-all generalization of binary classifiers) and the prediction of continuous outcomes.

Diffusion Monte Carlo approach versus adiabatic computation for local Hamiltonians

NASA Astrophysics Data System (ADS)

Bringewatt, Jacob; Dorland, William; Jordan, Stephen P.; Mink, Alan

2018-02-01

Most research regarding quantum adiabatic optimization has focused on stoquastic Hamiltonians, whose ground states can be expressed with only real non-negative amplitudes and thus for whom destructive interference is not manifest. This raises the question of whether classical Monte Carlo algorithms can efficiently simulate quantum adiabatic optimization with stoquastic Hamiltonians. Recent results have given counterexamples in which path-integral and diffusion Monte Carlo fail to do so. However, most adiabatic optimization algorithms, such as for solving MAX-k -SAT problems, use k -local Hamiltonians, whereas our previous counterexample for diffusion Monte Carlo involved n -body interactions. Here we present a 6-local counterexample which demonstrates that even for these local Hamiltonians there are cases where diffusion Monte Carlo cannot efficiently simulate quantum adiabatic optimization. Furthermore, we perform empirical testing of diffusion Monte Carlo on a standard well-studied class of permutation-symmetric tunneling problems and similarly find large advantages for quantum optimization over diffusion Monte Carlo.

The Impact of Monte Carlo Dose Calculations on Intensity-Modulated Radiation Therapy

NASA Astrophysics Data System (ADS)

Siebers, J. V.; Keall, P. J.; Mohan, R.

The effect of dose calculation accuracy for IMRT was studied by comparing different dose calculation algorithms. A head and neck IMRT plan was optimized using a superposition dose calculation algorithm. Dose was re-computed for the optimized plan using both Monte Carlo and pencil beam dose calculation algorithms to generate patient and phantom dose distributions. Tumor control probabilities (TCP) and normal tissue complication probabilities (NTCP) were computed to estimate the plan outcome. For the treatment plan studied, Monte Carlo best reproduces phantom dose measurements, the TCP was slightly lower than the superposition and pencil beam results, and the NTCP values differed little.

Neural network simulation of the atmospheric point spread function for the adjacency effect research

NASA Astrophysics Data System (ADS)

Ma, Xiaoshan; Wang, Haidong; Li, Ligang; Yang, Zhen; Meng, Xin

2016-10-01

Adjacency effect could be regarded as the convolution of the atmospheric point spread function (PSF) and the surface leaving radiance. Monte Carlo is a common method to simulate the atmospheric PSF. But it can't obtain analytic expression and the meaningful results can be only acquired by statistical analysis of millions of data. A backward Monte Carlo algorithm was employed to simulate photon emitting and propagating in the atmosphere under different conditions. The PSF was determined by recording the photon-receiving numbers in fixed bin at different position. A multilayer feed-forward neural network with a single hidden layer was designed to learn the relationship between the PSF's and the input condition parameters. The neural network used the back-propagation learning rule for training. Its input parameters involved atmosphere condition, spectrum range, observing geometry. The outputs of the network were photon-receiving numbers in the corresponding bin. Because the output units were too many to be allowed by neural network, the large network was divided into a collection of smaller ones. These small networks could be ran simultaneously on many workstations and/or PCs to speed up the training. It is important to note that the simulated PSF's by Monte Carlo technique in non-nadir viewing angles are more complicated than that in nadir conditions which brings difficulties in the design of the neural network. The results obtained show that the neural network approach could be very useful to compute the atmospheric PSF based on the simulated data generated by Monte Carlo method.

A Christoffel function weighted least squares algorithm for collocation approximations

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

Narayan, Akil; Jakeman, John D.; Zhou, Tao

Here, we propose, theoretically investigate, and numerically validate an algorithm for the Monte Carlo solution of least-squares polynomial approximation problems in a collocation framework. Our investigation is motivated by applications in the collocation approximation of parametric functions, which frequently entails construction of surrogates via orthogonal polynomials. A standard Monte Carlo approach would draw samples according to the density defining the orthogonal polynomial family. Our proposed algorithm instead samples with respect to the (weighted) pluripotential equilibrium measure of the domain, and subsequently solves a weighted least-squares problem, with weights given by evaluations of the Christoffel function. We present theoretical analysis tomore » motivate the algorithm, and numerical results that show our method is superior to standard Monte Carlo methods in many situations of interest.« less

A Christoffel function weighted least squares algorithm for collocation approximations

DOE PAGES

Narayan, Akil; Jakeman, John D.; Zhou, Tao

2016-11-28

Here, we propose, theoretically investigate, and numerically validate an algorithm for the Monte Carlo solution of least-squares polynomial approximation problems in a collocation framework. Our investigation is motivated by applications in the collocation approximation of parametric functions, which frequently entails construction of surrogates via orthogonal polynomials. A standard Monte Carlo approach would draw samples according to the density defining the orthogonal polynomial family. Our proposed algorithm instead samples with respect to the (weighted) pluripotential equilibrium measure of the domain, and subsequently solves a weighted least-squares problem, with weights given by evaluations of the Christoffel function. We present theoretical analysis tomore » motivate the algorithm, and numerical results that show our method is superior to standard Monte Carlo methods in many situations of interest.« less

Bayesian Analysis of High Dimensional Classification

NASA Astrophysics Data System (ADS)

Mukhopadhyay, Subhadeep; Liang, Faming

2009-12-01

Modern data mining and bioinformatics have presented an important playground for statistical learning techniques, where the number of input variables is possibly much larger than the sample size of the training data. In supervised learning, logistic regression or probit regression can be used to model a binary output and form perceptron classification rules based on Bayesian inference. In these cases , there is a lot of interest in searching for sparse model in High Dimensional regression(/classification) setup. we first discuss two common challenges for analyzing high dimensional data. The first one is the curse of dimensionality. The complexity of many existing algorithms scale exponentially with the dimensionality of the space and by virtue of that algorithms soon become computationally intractable and therefore inapplicable in many real applications. secondly, multicollinearities among the predictors which severely slowdown the algorithm. In order to make Bayesian analysis operational in high dimension we propose a novel 'Hierarchical stochastic approximation monte carlo algorithm' (HSAMC), which overcomes the curse of dimensionality, multicollinearity of predictors in high dimension and also it possesses the self-adjusting mechanism to avoid the local minima separated by high energy barriers. Models and methods are illustrated by simulation inspired from from the feild of genomics. Numerical results indicate that HSAMC can work as a general model selection sampler in high dimensional complex model space.

Introduction of Parallel GPGPU Acceleration Algorithms for the Solution of Radiative Transfer

NASA Technical Reports Server (NTRS)

Godoy, William F.; Liu, Xu

2011-01-01

General-purpose computing on graphics processing units (GPGPU) is a recent technique that allows the parallel graphics processing unit (GPU) to accelerate calculations performed sequentially by the central processing unit (CPU). To introduce GPGPU to radiative transfer, the Gauss-Seidel solution of the well-known expressions for 1-D and 3-D homogeneous, isotropic media is selected as a test case. Different algorithms are introduced to balance memory and GPU-CPU communication, critical aspects of GPGPU. Results show that speed-ups of one to two orders of magnitude are obtained when compared to sequential solutions. The underlying value of GPGPU is its potential extension in radiative solvers (e.g., Monte Carlo, discrete ordinates) at a minimal learning curve.

Monte Carlo evaluation of Acuros XB dose calculation Algorithm for intensity modulated radiation therapy of nasopharyngeal carcinoma

NASA Astrophysics Data System (ADS)

Yeh, Peter C. Y.; Lee, C. C.; Chao, T. C.; Tung, C. J.

2017-11-01

Intensity-modulated radiation therapy is an effective treatment modality for the nasopharyngeal carcinoma. One important aspect of this cancer treatment is the need to have an accurate dose algorithm dealing with the complex air/bone/tissue interface in the head-neck region to achieve the cure without radiation-induced toxicities. The Acuros XB algorithm explicitly solves the linear Boltzmann transport equation in voxelized volumes to account for the tissue heterogeneities such as lungs, bone, air, and soft tissues in the treatment field receiving radiotherapy. With the single beam setup in phantoms, this algorithm has already been demonstrated to achieve the comparable accuracy with Monte Carlo simulations. In the present study, five nasopharyngeal carcinoma patients treated with the intensity-modulated radiation therapy were examined for their dose distributions calculated using the Acuros XB in the planning target volume and the organ-at-risk. Corresponding results of Monte Carlo simulations were computed from the electronic portal image data and the BEAMnrc/DOSXYZnrc code. Analysis of dose distributions in terms of the clinical indices indicated that the Acuros XB was in comparable accuracy with Monte Carlo simulations and better than the anisotropic analytical algorithm for dose calculations in real patients.

Automatic mesh adaptivity for hybrid Monte Carlo/deterministic neutronics modeling of difficult shielding problems

DOE PAGES

Ibrahim, Ahmad M.; Wilson, Paul P.H.; Sawan, Mohamed E.; ...

2015-06-30

The CADIS and FW-CADIS hybrid Monte Carlo/deterministic techniques dramatically increase the efficiency of neutronics modeling, but their use in the accurate design analysis of very large and geometrically complex nuclear systems has been limited by the large number of processors and memory requirements for their preliminary deterministic calculations and final Monte Carlo calculation. Three mesh adaptivity algorithms were developed to reduce the memory requirements of CADIS and FW-CADIS without sacrificing their efficiency improvement. First, a macromaterial approach enhances the fidelity of the deterministic models without changing the mesh. Second, a deterministic mesh refinement algorithm generates meshes that capture as muchmore » geometric detail as possible without exceeding a specified maximum number of mesh elements. Finally, a weight window coarsening algorithm decouples the weight window mesh and energy bins from the mesh and energy group structure of the deterministic calculations in order to remove the memory constraint of the weight window map from the deterministic mesh resolution. The three algorithms were used to enhance an FW-CADIS calculation of the prompt dose rate throughout the ITER experimental facility. Using these algorithms resulted in a 23.3% increase in the number of mesh tally elements in which the dose rates were calculated in a 10-day Monte Carlo calculation and, additionally, increased the efficiency of the Monte Carlo simulation by a factor of at least 3.4. The three algorithms enabled this difficult calculation to be accurately solved using an FW-CADIS simulation on a regular computer cluster, eliminating the need for a world-class super computer.« less

Monte Carlo tests of the ELIPGRID-PC algorithm

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

Davidson, J.R.

1995-04-01

The standard tool for calculating the probability of detecting pockets of contamination called hot spots has been the ELIPGRID computer code of Singer and Wickman. The ELIPGRID-PC program has recently made this algorithm available for an IBM{reg_sign} PC. However, no known independent validation of the ELIPGRID algorithm exists. This document describes a Monte Carlo simulation-based validation of a modified version of the ELIPGRID-PC code. The modified ELIPGRID-PC code is shown to match Monte Carlo-calculated hot-spot detection probabilities to within {plus_minus}0.5% for 319 out of 320 test cases. The one exception, a very thin elliptical hot spot located within a rectangularmore » sampling grid, differed from the Monte Carlo-calculated probability by about 1%. These results provide confidence in the ability of the modified ELIPGRID-PC code to accurately predict hot-spot detection probabilities within an acceptable range of error.« less

A Monte Carlo Approach for Adaptive Testing with Content Constraints

ERIC Educational Resources Information Center

Belov, Dmitry I.; Armstrong, Ronald D.; Weissman, Alexander

2008-01-01

This article presents a new algorithm for computerized adaptive testing (CAT) when content constraints are present. The algorithm is based on shadow CAT methodology to meet content constraints but applies Monte Carlo methods and provides the following advantages over shadow CAT: (a) lower maximum item exposure rates, (b) higher utilization of the…

Teaching Markov Chain Monte Carlo: Revealing the Basic Ideas behind the Algorithm

ERIC Educational Resources Information Center

Stewart, Wayne; Stewart, Sepideh

2014-01-01

For many scientists, researchers and students Markov chain Monte Carlo (MCMC) simulation is an important and necessary tool to perform Bayesian analyses. The simulation is often presented as a mathematical algorithm and then translated into an appropriate computer program. However, this can result in overlooking the fundamental and deeper…

Finite element model updating using the shadow hybrid Monte Carlo technique

NASA Astrophysics Data System (ADS)

Boulkaibet, I.; Mthembu, L.; Marwala, T.; Friswell, M. I.; Adhikari, S.

2015-02-01

Recent research in the field of finite element model updating (FEM) advocates the adoption of Bayesian analysis techniques to dealing with the uncertainties associated with these models. However, Bayesian formulations require the evaluation of the Posterior Distribution Function which may not be available in analytical form. This is the case in FEM updating. In such cases sampling methods can provide good approximations of the Posterior distribution when implemented in the Bayesian context. Markov Chain Monte Carlo (MCMC) algorithms are the most popular sampling tools used to sample probability distributions. However, the efficiency of these algorithms is affected by the complexity of the systems (the size of the parameter space). The Hybrid Monte Carlo (HMC) offers a very important MCMC approach to dealing with higher-dimensional complex problems. The HMC uses the molecular dynamics (MD) steps as the global Monte Carlo (MC) moves to reach areas of high probability where the gradient of the log-density of the Posterior acts as a guide during the search process. However, the acceptance rate of HMC is sensitive to the system size as well as the time step used to evaluate the MD trajectory. To overcome this limitation we propose the use of the Shadow Hybrid Monte Carlo (SHMC) algorithm. The SHMC algorithm is a modified version of the Hybrid Monte Carlo (HMC) and designed to improve sampling for large-system sizes and time steps. This is done by sampling from a modified Hamiltonian function instead of the normal Hamiltonian function. In this paper, the efficiency and accuracy of the SHMC method is tested on the updating of two real structures; an unsymmetrical H-shaped beam structure and a GARTEUR SM-AG19 structure and is compared to the application of the HMC algorithm on the same structures.

Optimization of the Monte Carlo code for modeling of photon migration in tissue.

PubMed

Zołek, Norbert S; Liebert, Adam; Maniewski, Roman

2006-10-01

The Monte Carlo method is frequently used to simulate light transport in turbid media because of its simplicity and flexibility, allowing to analyze complicated geometrical structures. Monte Carlo simulations are, however, time consuming because of the necessity to track the paths of individual photons. The time consuming computation is mainly associated with the calculation of the logarithmic and trigonometric functions as well as the generation of pseudo-random numbers. In this paper, the Monte Carlo algorithm was developed and optimized, by approximation of the logarithmic and trigonometric functions. The approximations were based on polynomial and rational functions, and the errors of these approximations are less than 1% of the values of the original functions. The proposed algorithm was verified by simulations of the time-resolved reflectance at several source-detector separations. The results of the calculation using the approximated algorithm were compared with those of the Monte Carlo simulations obtained with an exact computation of the logarithm and trigonometric functions as well as with the solution of the diffusion equation. The errors of the moments of the simulated distributions of times of flight of photons (total number of photons, mean time of flight and variance) are less than 2% for a range of optical properties, typical of living tissues. The proposed approximated algorithm allows to speed up the Monte Carlo simulations by a factor of 4. The developed code can be used on parallel machines, allowing for further acceleration.

A High-Order Low-Order Algorithm with Exponentially Convergent Monte Carlo for Thermal Radiative Transfer

DOE PAGES

Bolding, Simon R.; Cleveland, Mathew Allen; Morel, Jim E.

2016-10-21

In this paper, we have implemented a new high-order low-order (HOLO) algorithm for solving thermal radiative transfer problems. The low-order (LO) system is based on the spatial and angular moments of the transport equation and a linear-discontinuous finite-element spatial representation, producing equations similar to the standard S 2 equations. The LO solver is fully implicit in time and efficiently resolves the nonlinear temperature dependence at each time step. The high-order (HO) solver utilizes exponentially convergent Monte Carlo (ECMC) to give a globally accurate solution for the angular intensity to a fixed-source pure-absorber transport problem. This global solution is used tomore » compute consistency terms, which require the HO and LO solutions to converge toward the same solution. The use of ECMC allows for the efficient reduction of statistical noise in the Monte Carlo solution, reducing inaccuracies introduced through the LO consistency terms. Finally, we compare results with an implicit Monte Carlo code for one-dimensional gray test problems and demonstrate the efficiency of ECMC over standard Monte Carlo in this HOLO algorithm.« less

Percentage depth dose evaluation in heterogeneous media using thermoluminescent dosimetry

PubMed Central

da Rosa, L.A.R.; Campos, L.T.; Alves, V.G.L.; Batista, D.V.S.; Facure, A.

2010-01-01

The purpose of this study is to investigate the influence of lung heterogeneity inside a soft tissue phantom on percentage depth dose (PDD). PDD curves were obtained experimentally using LiF:Mg,Ti (TLD‐100) thermoluminescent detectors and applying Eclipse treatment planning system algorithms Batho, modified Batho (M‐Batho or BMod), equivalent TAR (E‐TAR or EQTAR), and anisotropic analytical algorithm (AAA) for a 15 MV photon beam and field sizes of 1×1,2×2,5×5, and 10×10cm2. Monte Carlo simulations were performed using the DOSRZnrc user code of EGSnrc. The experimental results agree with Monte Carlo simulations for all irradiation field sizes. Comparisons with Monte Carlo calculations show that the AAA algorithm provides the best simulations of PDD curves for all field sizes investigated. However, even this algorithm cannot accurately predict PDD values in the lung for field sizes of 1×1 and 2×2cm2. An overdosage in the lung of about 40% and 20% is calculated by the AAA algorithm close to the interface soft tissue/lung for 1×1 and 2×2cm2 field sizes, respectively. It was demonstrated that differences of 100% between Monte Carlo results and the algorithms Batho, modified Batho, and equivalent TAR responses may exist inside the lung region for the 1×1cm2 field. PACS number: 87.55.kd

Random Numbers and Monte Carlo Methods

NASA Astrophysics Data System (ADS)

Scherer, Philipp O. J.

Many-body problems often involve the calculation of integrals of very high dimension which cannot be treated by standard methods. For the calculation of thermodynamic averages Monte Carlo methods are very useful which sample the integration volume at randomly chosen points. After summarizing some basic statistics, we discuss algorithms for the generation of pseudo-random numbers with given probability distribution which are essential for all Monte Carlo methods. We show how the efficiency of Monte Carlo integration can be improved by sampling preferentially the important configurations. Finally the famous Metropolis algorithm is applied to classical many-particle systems. Computer experiments visualize the central limit theorem and apply the Metropolis method to the traveling salesman problem.

Monte Carlo uncertainty analysis of dose estimates in radiochromic film dosimetry with single-channel and multichannel algorithms.

PubMed

Vera-Sánchez, Juan Antonio; Ruiz-Morales, Carmen; González-López, Antonio

2018-03-01

To provide a multi-stage model to calculate uncertainty in radiochromic film dosimetry with Monte-Carlo techniques. This new approach is applied to single-channel and multichannel algorithms. Two lots of Gafchromic EBT3 are exposed in two different Varian linacs. They are read with an EPSON V800 flatbed scanner. The Monte-Carlo techniques in uncertainty analysis provide a numerical representation of the probability density functions of the output magnitudes. From this numerical representation, traditional parameters of uncertainty analysis as the standard deviations and bias are calculated. Moreover, these numerical representations are used to investigate the shape of the probability density functions of the output magnitudes. Also, another calibration film is read in four EPSON scanners (two V800 and two 10000XL) and the uncertainty analysis is carried out with the four images. The dose estimates of single-channel and multichannel algorithms show a Gaussian behavior and low bias. The multichannel algorithms lead to less uncertainty in the final dose estimates when the EPSON V800 is employed as reading device. In the case of the EPSON 10000XL, the single-channel algorithms provide less uncertainty in the dose estimates for doses higher than four Gy. A multi-stage model has been presented. With the aid of this model and the use of the Monte-Carlo techniques, the uncertainty of dose estimates for single-channel and multichannel algorithms are estimated. The application of the model together with Monte-Carlo techniques leads to a complete characterization of the uncertainties in radiochromic film dosimetry. Copyright © 2018 Associazione Italiana di Fisica Medica. Published by Elsevier Ltd. All rights reserved.

CPMC-Lab: A MATLAB package for Constrained Path Monte Carlo calculations

NASA Astrophysics Data System (ADS)

Nguyen, Huy; Shi, Hao; Xu, Jie; Zhang, Shiwei

2014-12-01

We describe CPMC-Lab, a MATLAB program for the constrained-path and phaseless auxiliary-field Monte Carlo methods. These methods have allowed applications ranging from the study of strongly correlated models, such as the Hubbard model, to ab initio calculations in molecules and solids. The present package implements the full ground-state constrained-path Monte Carlo (CPMC) method in MATLAB with a graphical interface, using the Hubbard model as an example. The package can perform calculations in finite supercells in any dimensions, under periodic or twist boundary conditions. Importance sampling and all other algorithmic details of a total energy calculation are included and illustrated. This open-source tool allows users to experiment with various model and run parameters and visualize the results. It provides a direct and interactive environment to learn the method and study the code with minimal overhead for setup. Furthermore, the package can be easily generalized for auxiliary-field quantum Monte Carlo (AFQMC) calculations in many other models for correlated electron systems, and can serve as a template for developing a production code for AFQMC total energy calculations in real materials. Several illustrative studies are carried out in one- and two-dimensional lattices on total energy, kinetic energy, potential energy, and charge- and spin-gaps.

Experiences with Markov Chain Monte Carlo Convergence Assessment in Two Psychometric Examples

ERIC Educational Resources Information Center

Sinharay, Sandip

2004-01-01

There is an increasing use of Markov chain Monte Carlo (MCMC) algorithms for fitting statistical models in psychometrics, especially in situations where the traditional estimation techniques are very difficult to apply. One of the disadvantages of using an MCMC algorithm is that it is not straightforward to determine the convergence of the…

Metis: A Pure Metropolis Markov Chain Monte Carlo Bayesian Inference Library

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

Bates, Cameron Russell; Mckigney, Edward Allen

The use of Bayesian inference in data analysis has become the standard for large scienti c experiments [1, 2]. The Monte Carlo Codes Group(XCP-3) at Los Alamos has developed a simple set of algorithms currently implemented in C++ and Python to easily perform at-prior Markov Chain Monte Carlo Bayesian inference with pure Metropolis sampling. These implementations are designed to be user friendly and extensible for customization based on speci c application requirements. This document describes the algorithmic choices made and presents two use cases.

A histogram-free multicanonical Monte Carlo algorithm for the construction of analytical density of states

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

Eisenbach, Markus; Li, Ying Wai

We report a new multicanonical Monte Carlo (MC) algorithm to obtain the density of states (DOS) for physical systems with continuous state variables in statistical mechanics. Our algorithm is able to obtain an analytical form for the DOS expressed in a chosen basis set, instead of a numerical array of finite resolution as in previous variants of this class of MC methods such as the multicanonical (MUCA) sampling and Wang-Landau (WL) sampling. This is enabled by storing the visited states directly in a data set and avoiding the explicit collection of a histogram. This practice also has the advantage ofmore » avoiding undesirable artificial errors caused by the discretization and binning of continuous state variables. Our results show that this scheme is capable of obtaining converged results with a much reduced number of Monte Carlo steps, leading to a significant speedup over existing algorithms.« less

Decoding the non-stationary neuron spike trains by dual Monte Carlo point process estimation in motor Brain Machine Interfaces.

PubMed

Liao, Yuxi; Li, Hongbao; Zhang, Qiaosheng; Fan, Gong; Wang, Yiwen; Zheng, Xiaoxiang

2014-01-01

Decoding algorithm in motor Brain Machine Interfaces translates the neural signals to movement parameters. They usually assume the connection between the neural firings and movements to be stationary, which is not true according to the recent studies that observe the time-varying neuron tuning property. This property results from the neural plasticity and motor learning etc., which leads to the degeneration of the decoding performance when the model is fixed. To track the non-stationary neuron tuning during decoding, we propose a dual model approach based on Monte Carlo point process filtering method that enables the estimation also on the dynamic tuning parameters. When applied on both simulated neural signal and in vivo BMI data, the proposed adaptive method performs better than the one with static tuning parameters, which raises a promising way to design a long-term-performing model for Brain Machine Interfaces decoder.

Bayesian nonparametric dictionary learning for compressed sensing MRI.

PubMed

Huang, Yue; Paisley, John; Lin, Qin; Ding, Xinghao; Fu, Xueyang; Zhang, Xiao-Ping

2014-12-01

We develop a Bayesian nonparametric model for reconstructing magnetic resonance images (MRIs) from highly undersampled k -space data. We perform dictionary learning as part of the image reconstruction process. To this end, we use the beta process as a nonparametric dictionary learning prior for representing an image patch as a sparse combination of dictionary elements. The size of the dictionary and patch-specific sparsity pattern are inferred from the data, in addition to other dictionary learning variables. Dictionary learning is performed directly on the compressed image, and so is tailored to the MRI being considered. In addition, we investigate a total variation penalty term in combination with the dictionary learning model, and show how the denoising property of dictionary learning removes dependence on regularization parameters in the noisy setting. We derive a stochastic optimization algorithm based on Markov chain Monte Carlo for the Bayesian model, and use the alternating direction method of multipliers for efficiently performing total variation minimization. We present empirical results on several MRI, which show that the proposed regularization framework can improve reconstruction accuracy over other methods.

Forward and inverse uncertainty quantification using multilevel Monte Carlo algorithms for an elliptic non-local equation

DOE PAGES

Jasra, Ajay; Law, Kody J. H.; Zhou, Yan

2016-01-01

Our paper considers uncertainty quantification for an elliptic nonlocal equation. In particular, it is assumed that the parameters which define the kernel in the nonlocal operator are uncertain and a priori distributed according to a probability measure. It is shown that the induced probability measure on some quantities of interest arising from functionals of the solution to the equation with random inputs is well-defined,s as is the posterior distribution on parameters given observations. As the elliptic nonlocal equation cannot be solved approximate posteriors are constructed. The multilevel Monte Carlo (MLMC) and multilevel sequential Monte Carlo (MLSMC) sampling algorithms are usedmore » for a priori and a posteriori estimation, respectively, of quantities of interest. Furthermore, these algorithms reduce the amount of work to estimate posterior expectations, for a given level of error, relative to Monte Carlo and i.i.d. sampling from the posterior at a given level of approximation of the solution of the elliptic nonlocal equation.« less

Forward and inverse uncertainty quantification using multilevel Monte Carlo algorithms for an elliptic non-local equation

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

Jasra, Ajay; Law, Kody J. H.; Zhou, Yan

Our paper considers uncertainty quantification for an elliptic nonlocal equation. In particular, it is assumed that the parameters which define the kernel in the nonlocal operator are uncertain and a priori distributed according to a probability measure. It is shown that the induced probability measure on some quantities of interest arising from functionals of the solution to the equation with random inputs is well-defined,s as is the posterior distribution on parameters given observations. As the elliptic nonlocal equation cannot be solved approximate posteriors are constructed. The multilevel Monte Carlo (MLMC) and multilevel sequential Monte Carlo (MLSMC) sampling algorithms are usedmore » for a priori and a posteriori estimation, respectively, of quantities of interest. Furthermore, these algorithms reduce the amount of work to estimate posterior expectations, for a given level of error, relative to Monte Carlo and i.i.d. sampling from the posterior at a given level of approximation of the solution of the elliptic nonlocal equation.« less

Exact and Monte carlo resampling procedures for the Wilcoxon-Mann-Whitney and Kruskal-Wallis tests.

PubMed

Berry, K J; Mielke, P W

2000-12-01

Exact and Monte Carlo resampling FORTRAN programs are described for the Wilcoxon-Mann-Whitney rank sum test and the Kruskal-Wallis one-way analysis of variance for ranks test. The program algorithms compensate for tied values and do not depend on asymptotic approximations for probability values, unlike most algorithms contained in PC-based statistical software packages.

Pattern Recognition for a Flight Dynamics Monte Carlo Simulation

NASA Technical Reports Server (NTRS)

Restrepo, Carolina; Hurtado, John E.

2011-01-01

The design, analysis, and verification and validation of a spacecraft relies heavily on Monte Carlo simulations. Modern computational techniques are able to generate large amounts of Monte Carlo data but flight dynamics engineers lack the time and resources to analyze it all. The growing amounts of data combined with the diminished available time of engineers motivates the need to automate the analysis process. Pattern recognition algorithms are an innovative way of analyzing flight dynamics data efficiently. They can search large data sets for specific patterns and highlight critical variables so analysts can focus their analysis efforts. This work combines a few tractable pattern recognition algorithms with basic flight dynamics concepts to build a practical analysis tool for Monte Carlo simulations. Current results show that this tool can quickly and automatically identify individual design parameters, and most importantly, specific combinations of parameters that should be avoided in order to prevent specific system failures. The current version uses a kernel density estimation algorithm and a sequential feature selection algorithm combined with a k-nearest neighbor classifier to find and rank important design parameters. This provides an increased level of confidence in the analysis and saves a significant amount of time.

Unsupervised machine learning account of magnetic transitions in the Hubbard model

NASA Astrophysics Data System (ADS)

Ch'ng, Kelvin; Vazquez, Nick; Khatami, Ehsan

2018-01-01

We employ several unsupervised machine learning techniques, including autoencoders, random trees embedding, and t -distributed stochastic neighboring ensemble (t -SNE), to reduce the dimensionality of, and therefore classify, raw (auxiliary) spin configurations generated, through Monte Carlo simulations of small clusters, for the Ising and Fermi-Hubbard models at finite temperatures. Results from a convolutional autoencoder for the three-dimensional Ising model can be shown to produce the magnetization and the susceptibility as a function of temperature with a high degree of accuracy. Quantum fluctuations distort this picture and prevent us from making such connections between the output of the autoencoder and physical observables for the Hubbard model. However, we are able to define an indicator based on the output of the t -SNE algorithm that shows a near perfect agreement with the antiferromagnetic structure factor of the model in two and three spatial dimensions in the weak-coupling regime. t -SNE also predicts a transition to the canted antiferromagnetic phase for the three-dimensional model when a strong magnetic field is present. We show that these techniques cannot be expected to work away from half filling when the "sign problem" in quantum Monte Carlo simulations is present.

High order methods for the integration of the Bateman equations and other problems of the form of y‧ = F(y,t)y

NASA Astrophysics Data System (ADS)

Josey, C.; Forget, B.; Smith, K.

2017-12-01

This paper introduces two families of A-stable algorithms for the integration of y‧ = F (y , t) y: the extended predictor-corrector (EPC) and the exponential-linear (EL) methods. The structure of the algorithm families are described, and the method of derivation of the coefficients presented. The new algorithms are then tested on a simple deterministic problem and a Monte Carlo isotopic evolution problem. The EPC family is shown to be only second order for systems of ODEs. However, the EPC-RK45 algorithm had the highest accuracy on the Monte Carlo test, requiring at least a factor of 2 fewer function evaluations to achieve a given accuracy than a second order predictor-corrector method (center extrapolation / center midpoint method) with regards to Gd-157 concentration. Members of the EL family can be derived to at least fourth order. The EL3 and the EL4 algorithms presented are shown to be third and fourth order respectively on the systems of ODE test. In the Monte Carlo test, these methods did not overtake the accuracy of EPC methods before statistical uncertainty dominated the error. The statistical properties of the algorithms were also analyzed during the Monte Carlo problem. The new methods are shown to yield smaller standard deviations on final quantities as compared to the reference predictor-corrector method, by up to a factor of 1.4.

Rare Event Simulation in Radiation Transport

NASA Astrophysics Data System (ADS)

Kollman, Craig

This dissertation studies methods for estimating extremely small probabilities by Monte Carlo simulation. Problems in radiation transport typically involve estimating very rare events or the expected value of a random variable which is with overwhelming probability equal to zero. These problems often have high dimensional state spaces and irregular geometries so that analytic solutions are not possible. Monte Carlo simulation must be used to estimate the radiation dosage being transported to a particular location. If the area is well shielded the probability of any one particular particle getting through is very small. Because of the large number of particles involved, even a tiny fraction penetrating the shield may represent an unacceptable level of radiation. It therefore becomes critical to be able to accurately estimate this extremely small probability. Importance sampling is a well known technique for improving the efficiency of rare event calculations. Here, a new set of probabilities is used in the simulation runs. The results are multiplied by the likelihood ratio between the true and simulated probabilities so as to keep our estimator unbiased. The variance of the resulting estimator is very sensitive to which new set of transition probabilities are chosen. It is shown that a zero variance estimator does exist, but that its computation requires exact knowledge of the solution. A simple random walk with an associated killing model for the scatter of neutrons is introduced. Large deviation results for optimal importance sampling in random walks are extended to the case where killing is present. An adaptive "learning" algorithm for implementing importance sampling is given for more general Markov chain models of neutron scatter. For finite state spaces this algorithm is shown to give, with probability one, a sequence of estimates converging exponentially fast to the true solution. In the final chapter, an attempt to generalize this algorithm to a continuous state space is made. This involves partitioning the space into a finite number of cells. There is a tradeoff between additional computation per iteration and variance reduction per iteration that arises in determining the optimal grid size. All versions of this algorithm can be thought of as a compromise between deterministic and Monte Carlo methods, capturing advantages of both techniques.

GPU-powered Shotgun Stochastic Search for Dirichlet process mixtures of Gaussian Graphical Models

PubMed Central

Mukherjee, Chiranjit; Rodriguez, Abel

2016-01-01

Gaussian graphical models are popular for modeling high-dimensional multivariate data with sparse conditional dependencies. A mixture of Gaussian graphical models extends this model to the more realistic scenario where observations come from a heterogenous population composed of a small number of homogeneous sub-groups. In this paper we present a novel stochastic search algorithm for finding the posterior mode of high-dimensional Dirichlet process mixtures of decomposable Gaussian graphical models. Further, we investigate how to harness the massive thread-parallelization capabilities of graphical processing units to accelerate computation. The computational advantages of our algorithms are demonstrated with various simulated data examples in which we compare our stochastic search with a Markov chain Monte Carlo algorithm in moderate dimensional data examples. These experiments show that our stochastic search largely outperforms the Markov chain Monte Carlo algorithm in terms of computing-times and in terms of the quality of the posterior mode discovered. Finally, we analyze a gene expression dataset in which Markov chain Monte Carlo algorithms are too slow to be practically useful. PMID:28626348

GPU-powered Shotgun Stochastic Search for Dirichlet process mixtures of Gaussian Graphical Models.

PubMed

Mukherjee, Chiranjit; Rodriguez, Abel

2016-01-01

Gaussian graphical models are popular for modeling high-dimensional multivariate data with sparse conditional dependencies. A mixture of Gaussian graphical models extends this model to the more realistic scenario where observations come from a heterogenous population composed of a small number of homogeneous sub-groups. In this paper we present a novel stochastic search algorithm for finding the posterior mode of high-dimensional Dirichlet process mixtures of decomposable Gaussian graphical models. Further, we investigate how to harness the massive thread-parallelization capabilities of graphical processing units to accelerate computation. The computational advantages of our algorithms are demonstrated with various simulated data examples in which we compare our stochastic search with a Markov chain Monte Carlo algorithm in moderate dimensional data examples. These experiments show that our stochastic search largely outperforms the Markov chain Monte Carlo algorithm in terms of computing-times and in terms of the quality of the posterior mode discovered. Finally, we analyze a gene expression dataset in which Markov chain Monte Carlo algorithms are too slow to be practically useful.

Adaptive Sequential Monte Carlo for Multiple Changepoint Analysis

DOE PAGES

Heard, Nicholas A.; Turcotte, Melissa J. M.

2016-05-21

Process monitoring and control requires detection of structural changes in a data stream in real time. This paper introduces an efficient sequential Monte Carlo algorithm designed for learning unknown changepoints in continuous time. The method is intuitively simple: new changepoints for the latest window of data are proposed by conditioning only on data observed since the most recent estimated changepoint, as these observations carry most of the information about the current state of the process. The proposed method shows improved performance over the current state of the art. Another advantage of the proposed algorithm is that it can be mademore » adaptive, varying the number of particles according to the apparent local complexity of the target changepoint probability distribution. This saves valuable computing time when changes in the changepoint distribution are negligible, and enables re-balancing of the importance weights of existing particles when a significant change in the target distribution is encountered. The plain and adaptive versions of the method are illustrated using the canonical continuous time changepoint problem of inferring the intensity of an inhomogeneous Poisson process, although the method is generally applicable to any changepoint problem. Performance is demonstrated using both conjugate and non-conjugate Bayesian models for the intensity. Lastly, appendices to the article are available online, illustrating the method on other models and applications.« less

An Intentional Laboratory: The San Carlos Charter Learning Center.

ERIC Educational Resources Information Center

Darwish, Elise

2000-01-01

Describes the San Carlos Charter Learning Center, a K-8 school chartered by the San Carlos, California, school district to be a research and development site. It has successfully shared practices in multi-age groupings, interdisciplinary instruction, parents as teachers, and staff evaluation. The article expands on the school's challenges and…

Visual improvement for bad handwriting based on Monte-Carlo method

NASA Astrophysics Data System (ADS)

Shi, Cao; Xiao, Jianguo; Xu, Canhui; Jia, Wenhua

2014-03-01

A visual improvement algorithm based on Monte Carlo simulation is proposed in this paper, in order to enhance visual effects for bad handwriting. The whole improvement process is to use well designed typeface so as to optimize bad handwriting image. In this process, a series of linear operators for image transformation are defined for transforming typeface image to approach handwriting image. And specific parameters of linear operators are estimated by Monte Carlo method. Visual improvement experiments illustrate that the proposed algorithm can effectively enhance visual effect for handwriting image as well as maintain the original handwriting features, such as tilt, stroke order and drawing direction etc. The proposed visual improvement algorithm, in this paper, has a huge potential to be applied in tablet computer and Mobile Internet, in order to improve user experience on handwriting.

Information Theory, Inference and Learning Algorithms

NASA Astrophysics Data System (ADS)

Mackay, David J. C.

2003-10-01

Information theory and inference, often taught separately, are here united in one entertaining textbook. These topics lie at the heart of many exciting areas of contemporary science and engineering - communication, signal processing, data mining, machine learning, pattern recognition, computational neuroscience, bioinformatics, and cryptography. This textbook introduces theory in tandem with applications. Information theory is taught alongside practical communication systems, such as arithmetic coding for data compression and sparse-graph codes for error-correction. A toolbox of inference techniques, including message-passing algorithms, Monte Carlo methods, and variational approximations, are developed alongside applications of these tools to clustering, convolutional codes, independent component analysis, and neural networks. The final part of the book describes the state of the art in error-correcting codes, including low-density parity-check codes, turbo codes, and digital fountain codes -- the twenty-first century standards for satellite communications, disk drives, and data broadcast. Richly illustrated, filled with worked examples and over 400 exercises, some with detailed solutions, David MacKay's groundbreaking book is ideal for self-learning and for undergraduate or graduate courses. Interludes on crosswords, evolution, and sex provide entertainment along the way. In sum, this is a textbook on information, communication, and coding for a new generation of students, and an unparalleled entry point into these subjects for professionals in areas as diverse as computational biology, financial engineering, and machine learning.

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

Loading relativistic Maxwell distributions in particle simulations

NASA Astrophysics Data System (ADS)

Zenitani, S.

2015-12-01

In order to study energetic plasma phenomena by using particle-in-cell (PIC) and Monte-Carlo simulations, we need to deal with relativistic velocity distributions in these simulations. However, numerical algorithms to deal with relativistic distributions are not well known. In this contribution, we overview basic algorithms to load relativistic Maxwell distributions in PIC and Monte-Carlo simulations. For stationary relativistic Maxwellian, the inverse transform method and the Sobol algorithm are reviewed. To boost particles to obtain relativistic shifted-Maxwellian, two rejection methods are newly proposed in a physically transparent manner. Their acceptance efficiencies are 􏰅50% for generic cases and 100% for symmetric distributions. They can be combined with arbitrary base algorithms.

An unbiased Hessian representation for Monte Carlo PDFs.

PubMed

Carrazza, Stefano; Forte, Stefano; Kassabov, Zahari; Latorre, José Ignacio; Rojo, Juan

We develop a methodology for the construction of a Hessian representation of Monte Carlo sets of parton distributions, based on the use of a subset of the Monte Carlo PDF replicas as an unbiased linear basis, and of a genetic algorithm for the determination of the optimal basis. We validate the methodology by first showing that it faithfully reproduces a native Monte Carlo PDF set (NNPDF3.0), and then, that if applied to Hessian PDF set (MMHT14) which was transformed into a Monte Carlo set, it gives back the starting PDFs with minimal information loss. We then show that, when applied to a large Monte Carlo PDF set obtained as combination of several underlying sets, the methodology leads to a Hessian representation in terms of a rather smaller set of parameters (MC-H PDFs), thereby providing an alternative implementation of the recently suggested Meta-PDF idea and a Hessian version of the recently suggested PDF compression algorithm (CMC-PDFs). The mc2hessian conversion code is made publicly available together with (through LHAPDF6) a Hessian representations of the NNPDF3.0 set, and the MC-H PDF set.

SU-F-T-444: Quality Improvement Review of Radiation Therapy Treatment Planning in the Presence of Dental Implants

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

Parenica, H; Ford, J; Mavroidis, P

Purpose: To quantify and compare the effect of metallic dental implants (MDI) on dose distributions calculated using Collapsed Cone Convolution Superposition (CCCS) algorithm or a Monte Carlo algorithm (with and without correcting for the density of the MDI). Methods: Seven previously treated patients to the head and neck region were included in this study. The MDI and the streaking artifacts on the CT images were carefully contoured. For each patient a plan was optimized and calculated using the Pinnacle3 treatment planning system (TPS). For each patient two dose calculations were performed, a) with the densities of the MDI and CTmore » artifacts overridden (12 g/cc and 1 g/cc respectively) and b) without density overrides. The plans were then exported to the Monaco TPS and recalculated using Monte Carlo dose calculation algorithm. The changes in dose to PTVs and surrounding Regions of Interest (ROIs) were examined between all plans. Results: The Monte Carlo dose calculation indicated that PTVs received 6% lower dose than the CCCS algorithm predicted. In some cases, the Monte Carlo algorithm indicated that surrounding ROIs received higher dose (up to a factor of 2). Conclusion: Not properly accounting for dental implants can impact both the high dose regions (PTV) and the low dose regions (OAR). This study implies that if MDI and the artifacts are not appropriately contoured and given the correct density, there is potential significant impact on PTV coverage and OAR maximum doses.« less

Bayesian estimation of realized stochastic volatility model by Hybrid Monte Carlo algorithm

NASA Astrophysics Data System (ADS)

Takaishi, Tetsuya

2014-03-01

The hybrid Monte Carlo algorithm (HMCA) is applied for Bayesian parameter estimation of the realized stochastic volatility (RSV) model. Using the 2nd order minimum norm integrator (2MNI) for the molecular dynamics (MD) simulation in the HMCA, we find that the 2MNI is more efficient than the conventional leapfrog integrator. We also find that the autocorrelation time of the volatility variables sampled by the HMCA is very short. Thus it is concluded that the HMCA with the 2MNI is an efficient algorithm for parameter estimations of the RSV model.

Correction of Visual Perception Based on Neuro-Fuzzy Learning for the Humanoid Robot TEO.

PubMed

Hernandez-Vicen, Juan; Martinez, Santiago; Garcia-Haro, Juan Miguel; Balaguer, Carlos

2018-03-25

New applications related to robotic manipulation or transportation tasks, with or without physical grasping, are continuously being developed. To perform these activities, the robot takes advantage of different kinds of perceptions. One of the key perceptions in robotics is vision. However, some problems related to image processing makes the application of visual information within robot control algorithms difficult. Camera-based systems have inherent errors that affect the quality and reliability of the information obtained. The need of correcting image distortion slows down image parameter computing, which decreases performance of control algorithms. In this paper, a new approach to correcting several sources of visual distortions on images in only one computing step is proposed. The goal of this system/algorithm is the computation of the tilt angle of an object transported by a robot, minimizing image inherent errors and increasing computing speed. After capturing the image, the computer system extracts the angle using a Fuzzy filter that corrects at the same time all possible distortions, obtaining the real angle in only one processing step. This filter has been developed by the means of Neuro-Fuzzy learning techniques, using datasets with information obtained from real experiments. In this way, the computing time has been decreased and the performance of the application has been improved. The resulting algorithm has been tried out experimentally in robot transportation tasks in the humanoid robot TEO (Task Environment Operator) from the University Carlos III of Madrid.

Correction of Visual Perception Based on Neuro-Fuzzy Learning for the Humanoid Robot TEO

PubMed Central

2018-01-01

New applications related to robotic manipulation or transportation tasks, with or without physical grasping, are continuously being developed. To perform these activities, the robot takes advantage of different kinds of perceptions. One of the key perceptions in robotics is vision. However, some problems related to image processing makes the application of visual information within robot control algorithms difficult. Camera-based systems have inherent errors that affect the quality and reliability of the information obtained. The need of correcting image distortion slows down image parameter computing, which decreases performance of control algorithms. In this paper, a new approach to correcting several sources of visual distortions on images in only one computing step is proposed. The goal of this system/algorithm is the computation of the tilt angle of an object transported by a robot, minimizing image inherent errors and increasing computing speed. After capturing the image, the computer system extracts the angle using a Fuzzy filter that corrects at the same time all possible distortions, obtaining the real angle in only one processing step. This filter has been developed by the means of Neuro-Fuzzy learning techniques, using datasets with information obtained from real experiments. In this way, the computing time has been decreased and the performance of the application has been improved. The resulting algorithm has been tried out experimentally in robot transportation tasks in the humanoid robot TEO (Task Environment Operator) from the University Carlos III of Madrid. PMID:29587392

Testing trivializing maps in the Hybrid Monte Carlo algorithm

PubMed Central

Engel, Georg P.; Schaefer, Stefan

2011-01-01

We test a recent proposal to use approximate trivializing maps in a field theory to speed up Hybrid Monte Carlo simulations. Simulating the CPN−1 model, we find a small improvement with the leading order transformation, which is however compensated by the additional computational overhead. The scaling of the algorithm towards the continuum is not changed. In particular, the effect of the topological modes on the autocorrelation times is studied. PMID:21969733

Dynamic load balancing for petascale quantum Monte Carlo applications: The Alias method

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

Sudheer, C. D.; Krishnan, S.; Srinivasan, A.

Diffusion Monte Carlo is the most accurate widely used Quantum Monte Carlo method for the electronic structure of materials, but it requires frequent load balancing or population redistribution steps to maintain efficiency and avoid accumulation of systematic errors on parallel machines. The load balancing step can be a significant factor affecting performance, and will become more important as the number of processing elements increases. We propose a new dynamic load balancing algorithm, the Alias Method, and evaluate it theoretically and empirically. An important feature of the new algorithm is that the load can be perfectly balanced with each process receivingmore » at most one message. It is also optimal in the maximum size of messages received by any process. We also optimize its implementation to reduce network contention, a process facilitated by the low messaging requirement of the algorithm. Empirical results on the petaflop Cray XT Jaguar supercomputer at ORNL showing up to 30% improvement in performance on 120,000 cores. The load balancing algorithm may be straightforwardly implemented in existing codes. The algorithm may also be employed by any method with many near identical computational tasks that requires load balancing.« less

Stochastic evaluation of second-order many-body perturbation energies.

PubMed

Willow, Soohaeng Yoo; Kim, Kwang S; Hirata, So

2012-11-28

With the aid of the Laplace transform, the canonical expression of the second-order many-body perturbation correction to an electronic energy is converted into the sum of two 13-dimensional integrals, the 12-dimensional parts of which are evaluated by Monte Carlo integration. Weight functions are identified that are analytically normalizable, are finite and non-negative everywhere, and share the same singularities as the integrands. They thus generate appropriate distributions of four-electron walkers via the Metropolis algorithm, yielding correlation energies of small molecules within a few mE(h) of the correct values after 10(8) Monte Carlo steps. This algorithm does away with the integral transformation as the hotspot of the usual algorithms, has a far superior size dependence of cost, does not suffer from the sign problem of some quantum Monte Carlo methods, and potentially easily parallelizable and extensible to other more complex electron-correlation theories.

Hierarchical fractional-step approximations and parallel kinetic Monte Carlo algorithms

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

Arampatzis, Giorgos, E-mail: garab@math.uoc.gr; Katsoulakis, Markos A., E-mail: markos@math.umass.edu; Plechac, Petr, E-mail: plechac@math.udel.edu

2012-10-01

We present a mathematical framework for constructing and analyzing parallel algorithms for lattice kinetic Monte Carlo (KMC) simulations. The resulting algorithms have the capacity to simulate a wide range of spatio-temporal scales in spatially distributed, non-equilibrium physiochemical processes with complex chemistry and transport micro-mechanisms. Rather than focusing on constructing exactly the stochastic trajectories, our approach relies on approximating the evolution of observables, such as density, coverage, correlations and so on. More specifically, we develop a spatial domain decomposition of the Markov operator (generator) that describes the evolution of all observables according to the kinetic Monte Carlo algorithm. This domain decompositionmore » corresponds to a decomposition of the Markov generator into a hierarchy of operators and can be tailored to specific hierarchical parallel architectures such as multi-core processors or clusters of Graphical Processing Units (GPUs). Based on this operator decomposition, we formulate parallel Fractional step kinetic Monte Carlo algorithms by employing the Trotter Theorem and its randomized variants; these schemes, (a) are partially asynchronous on each fractional step time-window, and (b) are characterized by their communication schedule between processors. The proposed mathematical framework allows us to rigorously justify the numerical and statistical consistency of the proposed algorithms, showing the convergence of our approximating schemes to the original serial KMC. The approach also provides a systematic evaluation of different processor communicating schedules. We carry out a detailed benchmarking of the parallel KMC schemes using available exact solutions, for example, in Ising-type systems and we demonstrate the capabilities of the method to simulate complex spatially distributed reactions at very large scales on GPUs. Finally, we discuss work load balancing between processors and propose a re-balancing scheme based on probabilistic mass transport methods.« less

Annealed Importance Sampling Reversible Jump MCMC algorithms

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

Karagiannis, Georgios; Andrieu, Christophe

2013-03-20

It will soon be 20 years since reversible jump Markov chain Monte Carlo (RJ-MCMC) algorithms have been proposed. They have significantly extended the scope of Markov chain Monte Carlo simulation methods, offering the promise to be able to routinely tackle transdimensional sampling problems, as encountered in Bayesian model selection problems for example, in a principled and flexible fashion. Their practical efficient implementation, however, still remains a challenge. A particular difficulty encountered in practice is in the choice of the dimension matching variables (both their nature and their distribution) and the reversible transformations which allow one to define the one-to-one mappingsmore » underpinning the design of these algorithms. Indeed, even seemingly sensible choices can lead to algorithms with very poor performance. The focus of this paper is the development and performance evaluation of a method, annealed importance sampling RJ-MCMC (aisRJ), which addresses this problem by mitigating the sensitivity of RJ-MCMC algorithms to the aforementioned poor design. As we shall see the algorithm can be understood as being an “exact approximation” of an idealized MCMC algorithm that would sample from the model probabilities directly in a model selection set-up. Such an idealized algorithm may have good theoretical convergence properties, but typically cannot be implemented, and our algorithms can approximate the performance of such idealized algorithms to an arbitrary degree while not introducing any bias for any degree of approximation. Our approach combines the dimension matching ideas of RJ-MCMC with annealed importance sampling and its Markov chain Monte Carlo implementation. We illustrate the performance of the algorithm with numerical simulations which indicate that, although the approach may at first appear computationally involved, it is in fact competitive.« less

An improved target velocity sampling algorithm for free gas elastic scattering

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

Romano, Paul K.; Walsh, Jonathan A.

We present an improved algorithm for sampling the target velocity when simulating elastic scattering in a Monte Carlo neutron transport code that correctly accounts for the energy dependence of the scattering cross section. The algorithm samples the relative velocity directly, thereby avoiding a potentially inefficient rejection step based on the ratio of cross sections. Here, we have shown that this algorithm requires only one rejection step, whereas other methods of similar accuracy require two rejection steps. The method was verified against stochastic and deterministic reference results for upscattering percentages in 238U. Simulations of a light water reactor pin cell problemmore » demonstrate that using this algorithm results in a 3% or less penalty in performance when compared with an approximate method that is used in most production Monte Carlo codes« less

An improved target velocity sampling algorithm for free gas elastic scattering

DOE PAGES

Romano, Paul K.; Walsh, Jonathan A.

2018-02-03

We present an improved algorithm for sampling the target velocity when simulating elastic scattering in a Monte Carlo neutron transport code that correctly accounts for the energy dependence of the scattering cross section. The algorithm samples the relative velocity directly, thereby avoiding a potentially inefficient rejection step based on the ratio of cross sections. Here, we have shown that this algorithm requires only one rejection step, whereas other methods of similar accuracy require two rejection steps. The method was verified against stochastic and deterministic reference results for upscattering percentages in 238U. Simulations of a light water reactor pin cell problemmore » demonstrate that using this algorithm results in a 3% or less penalty in performance when compared with an approximate method that is used in most production Monte Carlo codes« less

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

Langan, Roisin T.; Archibald, Richard K.; Lamberti, Vincent

We have applied a new imputation-based method for analyzing incomplete data, called Monte Carlo Bayesian Database Generation (MCBDG), to the Spent Fuel Isotopic Composition (SFCOMPO) database. About 60% of the entries are absent for SFCOMPO. The method estimates missing values of a property from a probability distribution created from the existing data for the property, and then generates multiple instances of the completed database for training a machine learning algorithm. Uncertainty in the data is represented by an empirical or an assumed error distribution. The method makes few assumptions about the underlying data, and compares favorably against results obtained bymore » replacing missing information with constant values.« less

Simulation of Nuclear Reactor Kinetics by the Monte Carlo Method

NASA Astrophysics Data System (ADS)

Gomin, E. A.; Davidenko, V. D.; Zinchenko, A. S.; Kharchenko, I. K.

2017-12-01

The KIR computer code intended for calculations of nuclear reactor kinetics using the Monte Carlo method is described. The algorithm implemented in the code is described in detail. Some results of test calculations are given.

Markov chain Monte Carlo techniques applied to parton distribution functions determination: Proof of concept

NASA Astrophysics Data System (ADS)

Gbedo, Yémalin Gabin; Mangin-Brinet, Mariane

2017-07-01

We present a new procedure to determine parton distribution functions (PDFs), based on Markov chain Monte Carlo (MCMC) methods. The aim of this paper is to show that we can replace the standard χ2 minimization by procedures grounded on statistical methods, and on Bayesian inference in particular, thus offering additional insight into the rich field of PDFs determination. After a basic introduction to these techniques, we introduce the algorithm we have chosen to implement—namely Hybrid (or Hamiltonian) Monte Carlo. This algorithm, initially developed for Lattice QCD, turns out to be very interesting when applied to PDFs determination by global analyses; we show that it allows us to circumvent the difficulties due to the high dimensionality of the problem, in particular concerning the acceptance. A first feasibility study is performed and presented, which indicates that Markov chain Monte Carlo can successfully be applied to the extraction of PDFs and of their uncertainties.

Dose specification for radiation therapy: dose to water or dose to medium?

NASA Astrophysics Data System (ADS)

Ma, C.-M.; Li, Jinsheng

2011-05-01

The Monte Carlo method enables accurate dose calculation for radiation therapy treatment planning and has been implemented in some commercial treatment planning systems. Unlike conventional dose calculation algorithms that provide patient dose information in terms of dose to water with variable electron density, the Monte Carlo method calculates the energy deposition in different media and expresses dose to a medium. This paper discusses the differences in dose calculated using water with different electron densities and that calculated for different biological media and the clinical issues on dose specification including dose prescription and plan evaluation using dose to water and dose to medium. We will demonstrate that conventional photon dose calculation algorithms compute doses similar to those simulated by Monte Carlo using water with different electron densities, which are close (<4% differences) to doses to media but significantly different (up to 11%) from doses to water converted from doses to media following American Association of Physicists in Medicine (AAPM) Task Group 105 recommendations. Our results suggest that for consistency with previous radiation therapy experience Monte Carlo photon algorithms report dose to medium for radiotherapy dose prescription, treatment plan evaluation and treatment outcome analysis.

High-efficiency wavefunction updates for large scale Quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Kent, Paul; McDaniel, Tyler; Li, Ying Wai; D'Azevedo, Ed

Within ab intio Quantum Monte Carlo (QMC) simulations, the leading numerical cost for large systems is the computation of the values of the Slater determinants in the trial wavefunctions. The evaluation of each Monte Carlo move requires finding the determinant of a dense matrix, which is traditionally iteratively evaluated using a rank-1 Sherman-Morrison updating scheme to avoid repeated explicit calculation of the inverse. For calculations with thousands of electrons, this operation dominates the execution profile. We propose a novel rank- k delayed update scheme. This strategy enables probability evaluation for multiple successive Monte Carlo moves, with application of accepted moves to the matrices delayed until after a predetermined number of moves, k. Accepted events grouped in this manner are then applied to the matrices en bloc with enhanced arithmetic intensity and computational efficiency. This procedure does not change the underlying Monte Carlo sampling or the sampling efficiency. For large systems and algorithms such as diffusion Monte Carlo where the acceptance ratio is high, order of magnitude speedups can be obtained on both multi-core CPU and on GPUs, making this algorithm highly advantageous for current petascale and future exascale computations.

Quantifying the effect of air gap, depth, and range shifter thickness on TPS dosimetric accuracy in superficial PBS proton therapy.

PubMed

Shirey, Robert J; Wu, Hsinshun Terry

2018-01-01

This study quantifies the dosimetric accuracy of a commercial treatment planning system as functions of treatment depth, air gap, and range shifter thickness for superficial pencil beam scanning proton therapy treatments. The RayStation 6 pencil beam and Monte Carlo dose engines were each used to calculate the dose distributions for a single treatment plan with varying range shifter air gaps. Central axis dose values extracted from each of the calculated plans were compared to dose values measured with a calibrated PTW Markus chamber at various depths in RW3 solid water. Dose was measured at 12 depths, ranging from the surface to 5 cm, for each of the 18 different air gaps, which ranged from 0.5 to 28 cm. TPS dosimetric accuracy, defined as the ratio of calculated dose relative to the measured dose, was plotted as functions of depth and air gap for the pencil beam and Monte Carlo dose algorithms. The accuracy of the TPS pencil beam dose algorithm was found to be clinically unacceptable at depths shallower than 3 cm with air gaps wider than 10 cm, and increased range shifter thickness only added to the dosimetric inaccuracy of the pencil beam algorithm. Each configuration calculated with Monte Carlo was determined to be clinically acceptable. Further comparisons of the Monte Carlo dose algorithm to the measured spread-out Bragg Peaks of multiple fields used during machine commissioning verified the dosimetric accuracy of Monte Carlo in a variety of beam energies and field sizes. Discrepancies between measured and TPS calculated dose values can mainly be attributed to the ability (or lack thereof) of the TPS pencil beam dose algorithm to properly model secondary proton scatter generated in the range shifter. © 2017 The Authors. Journal of Applied Clinical Medical Physics published by Wiley Periodicals, Inc. on behalf of American Association of Physicists in Medicine.

Hybrid Microgrid Configuration Optimization with Evolutionary Algorithms

NASA Astrophysics Data System (ADS)

Lopez, Nicolas

This dissertation explores the Renewable Energy Integration Problem, and proposes a Genetic Algorithm embedded with a Monte Carlo simulation to solve large instances of the problem that are impractical to solve via full enumeration. The Renewable Energy Integration Problem is defined as finding the optimum set of components to supply the electric demand to a hybrid microgrid. The components considered are solar panels, wind turbines, diesel generators, electric batteries, connections to the power grid and converters, which can be inverters and/or rectifiers. The methodology developed is explained as well as the combinatorial formulation. In addition, 2 case studies of a single objective optimization version of the problem are presented, in order to minimize cost and to minimize global warming potential (GWP) followed by a multi-objective implementation of the offered methodology, by utilizing a non-sorting Genetic Algorithm embedded with a monte Carlo Simulation. The method is validated by solving a small instance of the problem with known solution via a full enumeration algorithm developed by NREL in their software HOMER. The dissertation concludes that the evolutionary algorithms embedded with Monte Carlo simulation namely modified Genetic Algorithms are an efficient form of solving the problem, by finding approximate solutions in the case of single objective optimization, and by approximating the true Pareto front in the case of multiple objective optimization of the Renewable Energy Integration Problem.

SU-F-T-148: Are the Approximations in Analytic Semi-Empirical Dose Calculation Algorithms for Intensity Modulated Proton Therapy for Complex Heterogeneities of Head and Neck Clinically Significant?

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

Yepes, P; UT MD Anderson Cancer Center, Houston, TX; Titt, U

2016-06-15

Purpose: Evaluate the differences in dose distributions between the proton analytic semi-empirical dose calculation algorithm used in the clinic and Monte Carlo calculations for a sample of 50 head-and-neck (H&N) patients and estimate the potential clinical significance of the differences. Methods: A cohort of 50 H&N patients, treated at the University of Texas Cancer Center with Intensity Modulated Proton Therapy (IMPT), were selected for evaluation of clinical significance of approximations in computed dose distributions. H&N site was selected because of the highly inhomogeneous nature of the anatomy. The Fast Dose Calculator (FDC), a fast track-repeating accelerated Monte Carlo algorithm formore » proton therapy, was utilized for the calculation of dose distributions delivered during treatment plans. Because of its short processing time, FDC allows for the processing of large cohorts of patients. FDC has been validated versus GEANT4, a full Monte Carlo system and measurements in water and for inhomogeneous phantoms. A gamma-index analysis, DVHs, EUDs, and TCP and NTCPs computed using published models were utilized to evaluate the differences between the Treatment Plan System (TPS) and FDC. Results: The Monte Carlo results systematically predict lower dose delivered in the target. The observed differences can be as large as 8 Gy, and should have a clinical impact. Gamma analysis also showed significant differences between both approaches, especially for the target volumes. Conclusion: Monte Carlo calculations with fast algorithms is practical and should be considered for the clinic, at least as a treatment plan verification tool.« less

The Metropolis Monte Carlo method with CUDA enabled Graphic Processing Units

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

Hall, Clifford; School of Physics, Astronomy, and Computational Sciences, George Mason University, 4400 University Dr., Fairfax, VA 22030; Ji, Weixiao

2014-02-01

We present a CPU–GPU system for runtime acceleration of large molecular simulations using GPU computation and memory swaps. The memory architecture of the GPU can be used both as container for simulation data stored on the graphics card and as floating-point code target, providing an effective means for the manipulation of atomistic or molecular data on the GPU. To fully take advantage of this mechanism, efficient GPU realizations of algorithms used to perform atomistic and molecular simulations are essential. Our system implements a versatile molecular engine, including inter-molecule interactions and orientational variables for performing the Metropolis Monte Carlo (MMC) algorithm,more » which is one type of Markov chain Monte Carlo. By combining memory objects with floating-point code fragments we have implemented an MMC parallel engine that entirely avoids the communication time of molecular data at runtime. Our runtime acceleration system is a forerunner of a new class of CPU–GPU algorithms exploiting memory concepts combined with threading for avoiding bus bandwidth and communication. The testbed molecular system used here is a condensed phase system of oligopyrrole chains. A benchmark shows a size scaling speedup of 60 for systems with 210,000 pyrrole monomers. Our implementation can easily be combined with MPI to connect in parallel several CPU–GPU duets. -- Highlights: •We parallelize the Metropolis Monte Carlo (MMC) algorithm on one CPU—GPU duet. •The Adaptive Tempering Monte Carlo employs MMC and profits from this CPU—GPU implementation. •Our benchmark shows a size scaling-up speedup of 62 for systems with 225,000 particles. •The testbed involves a polymeric system of oligopyrroles in the condensed phase. •The CPU—GPU parallelization includes dipole—dipole and Mie—Jones classic potentials.« less

A point kernel algorithm for microbeam radiation therapy

NASA Astrophysics Data System (ADS)

Debus, Charlotte; Oelfke, Uwe; Bartzsch, Stefan

2017-11-01

Microbeam radiation therapy (MRT) is a treatment approach in radiation therapy where the treatment field is spatially fractionated into arrays of a few tens of micrometre wide planar beams of unusually high peak doses separated by low dose regions of several hundred micrometre width. In preclinical studies, this treatment approach has proven to spare normal tissue more effectively than conventional radiation therapy, while being equally efficient in tumour control. So far dose calculations in MRT, a prerequisite for future clinical applications are based on Monte Carlo simulations. However, they are computationally expensive, since scoring volumes have to be small. In this article a kernel based dose calculation algorithm is presented that splits the calculation into photon and electron mediated energy transport, and performs the calculation of peak and valley doses in typical MRT treatment fields within a few minutes. Kernels are analytically calculated depending on the energy spectrum and material composition. In various homogeneous materials peak, valley doses and microbeam profiles are calculated and compared to Monte Carlo simulations. For a microbeam exposure of an anthropomorphic head phantom calculated dose values are compared to measurements and Monte Carlo calculations. Except for regions close to material interfaces calculated peak dose values match Monte Carlo results within 4% and valley dose values within 8% deviation. No significant differences are observed between profiles calculated by the kernel algorithm and Monte Carlo simulations. Measurements in the head phantom agree within 4% in the peak and within 10% in the valley region. The presented algorithm is attached to the treatment planning platform VIRTUOS. It was and is used for dose calculations in preclinical and pet-clinical trials at the biomedical beamline ID17 of the European synchrotron radiation facility in Grenoble, France.

Rational approximations to rational models: alternative algorithms for category learning.

PubMed

Sanborn, Adam N; Griffiths, Thomas L; Navarro, Daniel J

2010-10-01

Rational models of cognition typically consider the abstract computational problems posed by the environment, assuming that people are capable of optimally solving those problems. This differs from more traditional formal models of cognition, which focus on the psychological processes responsible for behavior. A basic challenge for rational models is thus explaining how optimal solutions can be approximated by psychological processes. We outline a general strategy for answering this question, namely to explore the psychological plausibility of approximation algorithms developed in computer science and statistics. In particular, we argue that Monte Carlo methods provide a source of rational process models that connect optimal solutions to psychological processes. We support this argument through a detailed example, applying this approach to Anderson's (1990, 1991) rational model of categorization (RMC), which involves a particularly challenging computational problem. Drawing on a connection between the RMC and ideas from nonparametric Bayesian statistics, we propose 2 alternative algorithms for approximate inference in this model. The algorithms we consider include Gibbs sampling, a procedure appropriate when all stimuli are presented simultaneously, and particle filters, which sequentially approximate the posterior distribution with a small number of samples that are updated as new data become available. Applying these algorithms to several existing datasets shows that a particle filter with a single particle provides a good description of human inferences.

Monte Carlo Simulations of Radiative and Neutrino Transport under Astrophysical Conditions

NASA Astrophysics Data System (ADS)

Krivosheyev, Yu. M.; Bisnovatyi-Kogan, G. S.

2018-05-01

Monte Carlo simulations are utilized to model radiative and neutrino transfer in astrophysics. An algorithm that can be used to study radiative transport in astrophysical plasma based on simulations of photon trajectories in a medium is described. Formation of the hard X-ray spectrum of the Galactic microquasar SS 433 is considered in detail as an example. Specific requirements for applying such simulations to neutrino transport in a densemedium and algorithmic differences compared to its application to photon transport are discussed.

Monte Carlo sampling of Wigner functions and surface hopping quantum dynamics

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

Kube, Susanna; Lasser, Caroline; Weber, Marcus

2009-04-01

The article addresses the achievable accuracy for a Monte Carlo sampling of Wigner functions in combination with a surface hopping algorithm for non-adiabatic quantum dynamics. The approximation of Wigner functions is realized by an adaption of the Metropolis algorithm for real-valued functions with disconnected support. The integration, which is necessary for computing values of the Wigner function, uses importance sampling with a Gaussian weight function. The numerical experiments agree with theoretical considerations and show an error of 2-3%.

An efficient Monte Carlo-based algorithm for scatter correction in keV cone-beam CT

NASA Astrophysics Data System (ADS)

Poludniowski, G.; Evans, P. M.; Hansen, V. N.; Webb, S.

2009-06-01

A new method is proposed for scatter-correction of cone-beam CT images. A coarse reconstruction is used in initial iteration steps. Modelling of the x-ray tube spectra and detector response are included in the algorithm. Photon diffusion inside the imaging subject is calculated using the Monte Carlo method. Photon scoring at the detector is calculated using forced detection to a fixed set of node points. The scatter profiles are then obtained by linear interpolation. The algorithm is referred to as the coarse reconstruction and fixed detection (CRFD) technique. Scatter predictions are quantitatively validated against a widely used general-purpose Monte Carlo code: BEAMnrc/EGSnrc (NRCC, Canada). Agreement is excellent. The CRFD algorithm was applied to projection data acquired with a Synergy XVI CBCT unit (Elekta Limited, Crawley, UK), using RANDO and Catphan phantoms (The Phantom Laboratory, Salem NY, USA). The algorithm was shown to be effective in removing scatter-induced artefacts from CBCT images, and took as little as 2 min on a desktop PC. Image uniformity was greatly improved as was CT-number accuracy in reconstructions. This latter improvement was less marked where the expected CT-number of a material was very different to the background material in which it was embedded.

Constant-pressure nested sampling with atomistic dynamics

NASA Astrophysics Data System (ADS)

Baldock, Robert J. N.; Bernstein, Noam; Salerno, K. Michael; Pártay, Lívia B.; Csányi, Gábor

2017-10-01

The nested sampling algorithm has been shown to be a general method for calculating the pressure-temperature-composition phase diagrams of materials. While the previous implementation used single-particle Monte Carlo moves, these are inefficient for condensed systems with general interactions where single-particle moves cannot be evaluated faster than the energy of the whole system. Here we enhance the method by using all-particle moves: either Galilean Monte Carlo or the total enthalpy Hamiltonian Monte Carlo algorithm, introduced in this paper. We show that these algorithms enable the determination of phase transition temperatures with equivalent accuracy to the previous method at 1 /N of the cost for an N -particle system with general interactions, or at equal cost when single-particle moves can be done in 1 /N of the cost of a full N -particle energy evaluation. We demonstrate this speed-up for the freezing and condensation transitions of the Lennard-Jones system and show the utility of the algorithms by calculating the order-disorder phase transition of a binary Lennard-Jones model alloy, the eutectic of copper-gold, the density anomaly of water, and the condensation and solidification of bead-spring polymers. The nested sampling method with all three algorithms is implemented in the pymatnest software.

A new moving strategy for the sequential Monte Carlo approach in optimizing the hydrological model parameters

NASA Astrophysics Data System (ADS)

Zhu, Gaofeng; Li, Xin; Ma, Jinzhu; Wang, Yunquan; Liu, Shaomin; Huang, Chunlin; Zhang, Kun; Hu, Xiaoli

2018-04-01

Sequential Monte Carlo (SMC) samplers have become increasing popular for estimating the posterior parameter distribution with the non-linear dependency structures and multiple modes often present in hydrological models. However, the explorative capabilities and efficiency of the sampler depends strongly on the efficiency in the move step of SMC sampler. In this paper we presented a new SMC sampler entitled the Particle Evolution Metropolis Sequential Monte Carlo (PEM-SMC) algorithm, which is well suited to handle unknown static parameters of hydrologic model. The PEM-SMC sampler is inspired by the works of Liang and Wong (2001) and operates by incorporating the strengths of the genetic algorithm, differential evolution algorithm and Metropolis-Hasting algorithm into the framework of SMC. We also prove that the sampler admits the target distribution to be a stationary distribution. Two case studies including a multi-dimensional bimodal normal distribution and a conceptual rainfall-runoff hydrologic model by only considering parameter uncertainty and simultaneously considering parameter and input uncertainty show that PEM-SMC sampler is generally superior to other popular SMC algorithms in handling the high dimensional problems. The study also indicated that it may be important to account for model structural uncertainty by using multiplier different hydrological models in the SMC framework in future study.

High-throughput Bayesian Network Learning using Heterogeneous Multicore Computers

PubMed Central

Linderman, Michael D.; Athalye, Vivek; Meng, Teresa H.; Asadi, Narges Bani; Bruggner, Robert; Nolan, Garry P.

2017-01-01

Aberrant intracellular signaling plays an important role in many diseases. The causal structure of signal transduction networks can be modeled as Bayesian Networks (BNs), and computationally learned from experimental data. However, learning the structure of Bayesian Networks (BNs) is an NP-hard problem that, even with fast heuristics, is too time consuming for large, clinically important networks (20–50 nodes). In this paper, we present a novel graphics processing unit (GPU)-accelerated implementation of a Monte Carlo Markov Chain-based algorithm for learning BNs that is up to 7.5-fold faster than current general-purpose processor (GPP)-based implementations. The GPU-based implementation is just one of several implementations within the larger application, each optimized for a different input or machine configuration. We describe the methodology we use to build an extensible application, assembled from these variants, that can target a broad range of heterogeneous systems, e.g., GPUs, multicore GPPs. Specifically we show how we use the Merge programming model to efficiently integrate, test and intelligently select among the different potential implementations. PMID:28819655

Hypothesis testing of scientific Monte Carlo calculations.

PubMed

Wallerberger, Markus; Gull, Emanuel

2017-11-01

The steadily increasing size of scientific Monte Carlo simulations and the desire for robust, correct, and reproducible results necessitates rigorous testing procedures for scientific simulations in order to detect numerical problems and programming bugs. However, the testing paradigms developed for deterministic algorithms have proven to be ill suited for stochastic algorithms. In this paper we demonstrate explicitly how the technique of statistical hypothesis testing, which is in wide use in other fields of science, can be used to devise automatic and reliable tests for Monte Carlo methods, and we show that these tests are able to detect some of the common problems encountered in stochastic scientific simulations. We argue that hypothesis testing should become part of the standard testing toolkit for scientific simulations.

Hypothesis testing of scientific Monte Carlo calculations

NASA Astrophysics Data System (ADS)

Wallerberger, Markus; Gull, Emanuel

2017-11-01

The steadily increasing size of scientific Monte Carlo simulations and the desire for robust, correct, and reproducible results necessitates rigorous testing procedures for scientific simulations in order to detect numerical problems and programming bugs. However, the testing paradigms developed for deterministic algorithms have proven to be ill suited for stochastic algorithms. In this paper we demonstrate explicitly how the technique of statistical hypothesis testing, which is in wide use in other fields of science, can be used to devise automatic and reliable tests for Monte Carlo methods, and we show that these tests are able to detect some of the common problems encountered in stochastic scientific simulations. We argue that hypothesis testing should become part of the standard testing toolkit for scientific simulations.

Online sequential Monte Carlo smoother for partially observed diffusion processes

NASA Astrophysics Data System (ADS)

Gloaguen, Pierre; Étienne, Marie-Pierre; Le Corff, Sylvain

2018-12-01

This paper introduces a new algorithm to approximate smoothed additive functionals of partially observed diffusion processes. This method relies on a new sequential Monte Carlo method which allows to compute such approximations online, i.e., as the observations are received, and with a computational complexity growing linearly with the number of Monte Carlo samples. The original algorithm cannot be used in the case of partially observed stochastic differential equations since the transition density of the latent data is usually unknown. We prove that it may be extended to partially observed continuous processes by replacing this unknown quantity by an unbiased estimator obtained for instance using general Poisson estimators. This estimator is proved to be consistent and its performance are illustrated using data from two models.

A surrogate accelerated multicanonical Monte Carlo method for uncertainty quantification

NASA Astrophysics Data System (ADS)

Wu, Keyi; Li, Jinglai

2016-09-01

In this work we consider a class of uncertainty quantification problems where the system performance or reliability is characterized by a scalar parameter y. The performance parameter y is random due to the presence of various sources of uncertainty in the system, and our goal is to estimate the probability density function (PDF) of y. We propose to use the multicanonical Monte Carlo (MMC) method, a special type of adaptive importance sampling algorithms, to compute the PDF of interest. Moreover, we develop an adaptive algorithm to construct local Gaussian process surrogates to further accelerate the MMC iterations. With numerical examples we demonstrate that the proposed method can achieve several orders of magnitudes of speedup over the standard Monte Carlo methods.

Path integral Monte Carlo ground state approach: formalism, implementation, and applications

NASA Astrophysics Data System (ADS)

Yan, Yangqian; Blume, D.

2017-11-01

Monte Carlo techniques have played an important role in understanding strongly correlated systems across many areas of physics, covering a wide range of energy and length scales. Among the many Monte Carlo methods applicable to quantum mechanical systems, the path integral Monte Carlo approach with its variants has been employed widely. Since semi-classical or classical approaches will not be discussed in this review, path integral based approaches can for our purposes be divided into two categories: approaches applicable to quantum mechanical systems at zero temperature and approaches applicable to quantum mechanical systems at finite temperature. While these two approaches are related to each other, the underlying formulation and aspects of the algorithm differ. This paper reviews the path integral Monte Carlo ground state (PIGS) approach, which solves the time-independent Schrödinger equation. Specifically, the PIGS approach allows for the determination of expectation values with respect to eigen states of the few- or many-body Schrödinger equation provided the system Hamiltonian is known. The theoretical framework behind the PIGS algorithm, implementation details, and sample applications for fermionic systems are presented.

Multigroup Monte Carlo on GPUs: Comparison of history- and event-based algorithms

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

Hamilton, Steven P.; Slattery, Stuart R.; Evans, Thomas M.

This article presents an investigation of the performance of different multigroup Monte Carlo transport algorithms on GPUs with a discussion of both history-based and event-based approaches. Several algorithmic improvements are introduced for both approaches. By modifying the history-based algorithm that is traditionally favored in CPU-based MC codes to occasionally filter out dead particles to reduce thread divergence, performance exceeds that of either the pure history-based or event-based approaches. The impacts of several algorithmic choices are discussed, including performance studies on Kepler and Pascal generation NVIDIA GPUs for fixed source and eigenvalue calculations. Single-device performance equivalent to 20–40 CPU cores onmore » the K40 GPU and 60–80 CPU cores on the P100 GPU is achieved. Last, in addition, nearly perfect multi-device parallel weak scaling is demonstrated on more than 16,000 nodes of the Titan supercomputer.« less

Multigroup Monte Carlo on GPUs: Comparison of history- and event-based algorithms

DOE PAGES

Hamilton, Steven P.; Slattery, Stuart R.; Evans, Thomas M.

2017-12-22

This article presents an investigation of the performance of different multigroup Monte Carlo transport algorithms on GPUs with a discussion of both history-based and event-based approaches. Several algorithmic improvements are introduced for both approaches. By modifying the history-based algorithm that is traditionally favored in CPU-based MC codes to occasionally filter out dead particles to reduce thread divergence, performance exceeds that of either the pure history-based or event-based approaches. The impacts of several algorithmic choices are discussed, including performance studies on Kepler and Pascal generation NVIDIA GPUs for fixed source and eigenvalue calculations. Single-device performance equivalent to 20–40 CPU cores onmore » the K40 GPU and 60–80 CPU cores on the P100 GPU is achieved. Last, in addition, nearly perfect multi-device parallel weak scaling is demonstrated on more than 16,000 nodes of the Titan supercomputer.« less

Comments on "Including the effects of temperature-dependent opacities in the implicit Monte Carlo algorithm" by N.A. Gentile [J. Comput. Phys. 230 (2011) 5100-5114

NASA Astrophysics Data System (ADS)

Ghosh, Karabi

2017-02-01

We briefly comment on a paper by N.A. Gentile [J. Comput. Phys. 230 (2011) 5100-5114] in which the Fleck factor has been modified to include the effects of temperature-dependent opacities in the implicit Monte Carlo algorithm developed by Fleck and Cummings [1,2]. Instead of the Fleck factor, f = 1 / (1 + βcΔtσP), the author derived the modified Fleck factor g = 1 / (1 + βcΔtσP - min [σP‧ (aTr4 - aT4)cΔt/ρCV, 0 ]) to be used in the Implicit Monte Carlo (IMC) algorithm in order to obtain more accurate solutions with much larger time steps. Here β = 4 aT3 / ρCV, σP is the Planck opacity and the derivative of Planck opacity w.r.t. the material temperature is σP‧ = dσP / dT.

Parallel Algorithms for Monte Carlo Particle Transport Simulation on Exascale Computing Architectures

NASA Astrophysics Data System (ADS)

Romano, Paul Kollath

Monte Carlo particle transport methods are being considered as a viable option for high-fidelity simulation of nuclear reactors. While Monte Carlo methods offer several potential advantages over deterministic methods, there are a number of algorithmic shortcomings that would prevent their immediate adoption for full-core analyses. In this thesis, algorithms are proposed both to ameliorate the degradation in parallel efficiency typically observed for large numbers of processors and to offer a means of decomposing large tally data that will be needed for reactor analysis. A nearest-neighbor fission bank algorithm was proposed and subsequently implemented in the OpenMC Monte Carlo code. A theoretical analysis of the communication pattern shows that the expected cost is O( N ) whereas traditional fission bank algorithms are O(N) at best. The algorithm was tested on two supercomputers, the Intrepid Blue Gene/P and the Titan Cray XK7, and demonstrated nearly linear parallel scaling up to 163,840 processor cores on a full-core benchmark problem. An algorithm for reducing network communication arising from tally reduction was analyzed and implemented in OpenMC. The proposed algorithm groups only particle histories on a single processor into batches for tally purposes---in doing so it prevents all network communication for tallies until the very end of the simulation. The algorithm was tested, again on a full-core benchmark, and shown to reduce network communication substantially. A model was developed to predict the impact of load imbalances on the performance of domain decomposed simulations. The analysis demonstrated that load imbalances in domain decomposed simulations arise from two distinct phenomena: non-uniform particle densities and non-uniform spatial leakage. The dominant performance penalty for domain decomposition was shown to come from these physical effects rather than insufficient network bandwidth or high latency. The model predictions were verified with measured data from simulations in OpenMC on a full-core benchmark problem. Finally, a novel algorithm for decomposing large tally data was proposed, analyzed, and implemented/tested in OpenMC. The algorithm relies on disjoint sets of compute processes and tally servers. The analysis showed that for a range of parameters relevant to LWR analysis, the tally server algorithm should perform with minimal overhead. Tests were performed on Intrepid and Titan and demonstrated that the algorithm did indeed perform well over a wide range of parameters. (Copies available exclusively from MIT Libraries, libraries.mit.edu/docs - docs mit.edu)

Supervised Learning of Two-Layer Perceptron under the Existence of External Noise — Learning Curve of Boolean Functions of Two Variables in Tree-Like Architecture —

NASA Astrophysics Data System (ADS)

Uezu, Tatsuya; Kiyokawa, Shuji

2016-06-01

We investigate the supervised batch learning of Boolean functions expressed by a two-layer perceptron with a tree-like structure. We adopt continuous weights (spherical model) and the Gibbs algorithm. We study the Parity and And machines and two types of noise, input and output noise, together with the noiseless case. We assume that only the teacher suffers from noise. By using the replica method, we derive the saddle point equations for order parameters under the replica symmetric (RS) ansatz. We study the critical value αC of the loading rate α above which the learning phase exists for cases with and without noise. We find that αC is nonzero for the Parity machine, while it is zero for the And machine. We derive the exponents barβ of order parameters expressed as (α - α C)bar{β} when α is near to αC. Furthermore, in the Parity machine, when noise exists, we find a spin glass solution, in which the overlap between the teacher and student vectors is zero but that between student vectors is nonzero. We perform Markov chain Monte Carlo simulations by simulated annealing and also by exchange Monte Carlo simulations in both machines. In the Parity machine, we study the de Almeida-Thouless stability, and by comparing theoretical and numerical results, we find that there exist parameter regions where the RS solution is unstable, and that the spin glass solution is metastable or unstable. We also study asymptotic learning behavior for large α and derive the exponents hat{β } of order parameters expressed as α - hat{β } when α is large in both machines. By simulated annealing simulations, we confirm these results and conclude that learning takes place for the input noise case with any noise amplitude and for the output noise case when the probability that the teacher's output is reversed is less than one-half.

MDTS: automatic complex materials design using Monte Carlo tree search.

PubMed

M Dieb, Thaer; Ju, Shenghong; Yoshizoe, Kazuki; Hou, Zhufeng; Shiomi, Junichiro; Tsuda, Koji

2017-01-01

Complex materials design is often represented as a black-box combinatorial optimization problem. In this paper, we present a novel python library called MDTS (Materials Design using Tree Search). Our algorithm employs a Monte Carlo tree search approach, which has shown exceptional performance in computer Go game. Unlike evolutionary algorithms that require user intervention to set parameters appropriately, MDTS has no tuning parameters and works autonomously in various problems. In comparison to a Bayesian optimization package, our algorithm showed competitive search efficiency and superior scalability. We succeeded in designing large Silicon-Germanium (Si-Ge) alloy structures that Bayesian optimization could not deal with due to excessive computational cost. MDTS is available at https://github.com/tsudalab/MDTS.

MDTS: automatic complex materials design using Monte Carlo tree search

NASA Astrophysics Data System (ADS)

Dieb, Thaer M.; Ju, Shenghong; Yoshizoe, Kazuki; Hou, Zhufeng; Shiomi, Junichiro; Tsuda, Koji

2017-12-01

Complex materials design is often represented as a black-box combinatorial optimization problem. In this paper, we present a novel python library called MDTS (Materials Design using Tree Search). Our algorithm employs a Monte Carlo tree search approach, which has shown exceptional performance in computer Go game. Unlike evolutionary algorithms that require user intervention to set parameters appropriately, MDTS has no tuning parameters and works autonomously in various problems. In comparison to a Bayesian optimization package, our algorithm showed competitive search efficiency and superior scalability. We succeeded in designing large Silicon-Germanium (Si-Ge) alloy structures that Bayesian optimization could not deal with due to excessive computational cost. MDTS is available at https://github.com/tsudalab/MDTS.

Efficient kinetic Monte Carlo method for reaction-diffusion problems with spatially varying annihilation rates

NASA Astrophysics Data System (ADS)

Schwarz, Karsten; Rieger, Heiko

2013-03-01

We present an efficient Monte Carlo method to simulate reaction-diffusion processes with spatially varying particle annihilation or transformation rates as it occurs for instance in the context of motor-driven intracellular transport. Like Green's function reaction dynamics and first-passage time methods, our algorithm avoids small diffusive hops by propagating sufficiently distant particles in large hops to the boundaries of protective domains. Since for spatially varying annihilation or transformation rates the single particle diffusion propagator is not known analytically, we present an algorithm that generates efficiently either particle displacements or annihilations with the correct statistics, as we prove rigorously. The numerical efficiency of the algorithm is demonstrated with an illustrative example.

Note: A pure-sampling quantum Monte Carlo algorithm with independent Metropolis.

PubMed

Vrbik, Jan; Ospadov, Egor; Rothstein, Stuart M

2016-07-14

Recently, Ospadov and Rothstein published a pure-sampling quantum Monte Carlo algorithm (PSQMC) that features an auxiliary Path Z that connects the midpoints of the current and proposed Paths X and Y, respectively. When sufficiently long, Path Z provides statistical independence of Paths X and Y. Under those conditions, the Metropolis decision used in PSQMC is done without any approximation, i.e., not requiring microscopic reversibility and without having to introduce any G(x → x'; τ) factors into its decision function. This is a unique feature that contrasts with all competing reptation algorithms in the literature. An example illustrates that dependence of Paths X and Y has adverse consequences for pure sampling.

Note: A pure-sampling quantum Monte Carlo algorithm with independent Metropolis

NASA Astrophysics Data System (ADS)

Vrbik, Jan; Ospadov, Egor; Rothstein, Stuart M.

2016-07-01

Recently, Ospadov and Rothstein published a pure-sampling quantum Monte Carlo algorithm (PSQMC) that features an auxiliary Path Z that connects the midpoints of the current and proposed Paths X and Y, respectively. When sufficiently long, Path Z provides statistical independence of Paths X and Y. Under those conditions, the Metropolis decision used in PSQMC is done without any approximation, i.e., not requiring microscopic reversibility and without having to introduce any G(x → x'; τ) factors into its decision function. This is a unique feature that contrasts with all competing reptation algorithms in the literature. An example illustrates that dependence of Paths X and Y has adverse consequences for pure sampling.

Event-driven Monte Carlo: Exact dynamics at all time scales for discrete-variable models

NASA Astrophysics Data System (ADS)

Mendoza-Coto, Alejandro; Díaz-Méndez, Rogelio; Pupillo, Guido

2016-06-01

We present an algorithm for the simulation of the exact real-time dynamics of classical many-body systems with discrete energy levels. In the same spirit of kinetic Monte Carlo methods, a stochastic solution of the master equation is found, with no need to define any other phase-space construction. However, unlike existing methods, the present algorithm does not assume any particular statistical distribution to perform moves or to advance the time, and thus is a unique tool for the numerical exploration of fast and ultra-fast dynamical regimes. By decomposing the problem in a set of two-level subsystems, we find a natural variable step size, that is well defined from the normalization condition of the transition probabilities between the levels. We successfully test the algorithm with known exact solutions for non-equilibrium dynamics and equilibrium thermodynamical properties of Ising-spin models in one and two dimensions, and compare to standard implementations of kinetic Monte Carlo methods. The present algorithm is directly applicable to the study of the real-time dynamics of a large class of classical Markovian chains, and particularly to short-time situations where the exact evolution is relevant.

Geometrically Constructed Markov Chain Monte Carlo Study of Quantum Spin-phonon Complex Systems

NASA Astrophysics Data System (ADS)

Suwa, Hidemaro

2013-03-01

We have developed novel Monte Carlo methods for precisely calculating quantum spin-boson models and investigated the critical phenomena of the spin-Peierls systems. Three significant methods are presented. The first is a new optimization algorithm of the Markov chain transition kernel based on the geometric weight allocation. This algorithm, for the first time, satisfies the total balance generally without imposing the detailed balance and always minimizes the average rejection rate, being better than the Metropolis algorithm. The second is the extension of the worm (directed-loop) algorithm to non-conserved particles, which cannot be treated efficiently by the conventional methods. The third is the combination with the level spectroscopy. Proposing a new gap estimator, we are successful in eliminating the systematic error of the conventional moment method. Then we have elucidated the phase diagram and the universality class of the one-dimensional XXZ spin-Peierls system. The criticality is totally consistent with the J1 -J2 model, an effective model in the antiadiabatic limit. Through this research, we have succeeded in investigating the critical phenomena of the effectively frustrated quantum spin system by the quantum Monte Carlo method without the negative sign. JSPS Postdoctoral Fellow for Research Abroad

Technical Note: A direct ray-tracing method to compute integral depth dose in pencil beam proton radiography with a multilayer ionization chamber.

PubMed

Farace, Paolo; Righetto, Roberto; Deffet, Sylvain; Meijers, Arturs; Vander Stappen, Francois

2016-12-01

To introduce a fast ray-tracing algorithm in pencil proton radiography (PR) with a multilayer ionization chamber (MLIC) for in vivo range error mapping. Pencil beam PR was obtained by delivering spots uniformly positioned in a square (45 × 45 mm 2 field-of-view) of 9 × 9 spots capable of crossing the phantoms (210 MeV). The exit beam was collected by a MLIC to sample the integral depth dose (IDD MLIC ). PRs of an electron-density and of a head phantom were acquired by moving the couch to obtain multiple 45 × 45 mm 2 frames. To map the corresponding range errors, the two-dimensional set of IDD MLIC was compared with (i) the integral depth dose computed by the treatment planning system (TPS) by both analytic (IDD TPS ) and Monte Carlo (IDD MC ) algorithms in a volume of water simulating the MLIC at the CT, and (ii) the integral depth dose directly computed by a simple ray-tracing algorithm (IDD direct ) through the same CT data. The exact spatial position of the spot pattern was numerically adjusted testing different in-plane positions and selecting the one that minimized the range differences between IDD direct and IDD MLIC . Range error mapping was feasible by both the TPS and the ray-tracing methods, but very sensitive to even small misalignments. In homogeneous regions, the range errors computed by the direct ray-tracing algorithm matched the results obtained by both the analytic and the Monte Carlo algorithms. In both phantoms, lateral heterogeneities were better modeled by the ray-tracing and the Monte Carlo algorithms than by the analytic TPS computation. Accordingly, when the pencil beam crossed lateral heterogeneities, the range errors mapped by the direct algorithm matched better the Monte Carlo maps than those obtained by the analytic algorithm. Finally, the simplicity of the ray-tracing algorithm allowed to implement a prototype procedure for automated spatial alignment. The ray-tracing algorithm can reliably replace the TPS method in MLIC PR for in vivo range verification and it can be a key component to develop software tools for spatial alignment and correction of CT calibration.

A robust algorithm for automated target recognition using precomputed radar cross sections

NASA Astrophysics Data System (ADS)

Ehrman, Lisa M.; Lanterman, Aaron D.

2004-09-01

Passive radar is an emerging technology that offers a number of unique benefits, including covert operation. Many such systems are already capable of detecting and tracking aircraft. The goal of this work is to develop a robust algorithm for adding automated target recognition (ATR) capabilities to existing passive radar systems. In previous papers, we proposed conducting ATR by comparing the precomputed RCS of known targets to that of detected targets. To make the precomputed RCS as accurate as possible, a coordinated flight model is used to estimate aircraft orientation. Once the aircraft's position and orientation are known, it is possible to determine the incident and observed angles on the aircraft, relative to the transmitter and receiver. This makes it possible to extract the appropriate radar cross section (RCS) from our simulated database. This RCS is then scaled to account for propagation losses and the receiver's antenna gain. A Rician likelihood model compares these expected signals from different targets to the received target profile. We have previously employed Monte Carlo runs to gauge the probability of error in the ATR algorithm; however, generation of a statistically significant set of Monte Carlo runs is computationally intensive. As an alternative to Monte Carlo runs, we derive the relative entropy (also known as Kullback-Liebler distance) between two Rician distributions. Since the probability of Type II error in our hypothesis testing problem can be expressed as a function of the relative entropy via Stein's Lemma, this provides us with a computationally efficient method for determining an upper bound on our algorithm's performance. It also provides great insight into the types of classification errors we can expect from our algorithm. This paper compares the numerically approximated probability of Type II error with the results obtained from a set of Monte Carlo runs.

Monte Carlo reference data sets for imaging research: Executive summary of the report of AAPM Research Committee Task Group 195.

PubMed

Sechopoulos, Ioannis; Ali, Elsayed S M; Badal, Andreu; Badano, Aldo; Boone, John M; Kyprianou, Iacovos S; Mainegra-Hing, Ernesto; McMillan, Kyle L; McNitt-Gray, Michael F; Rogers, D W O; Samei, Ehsan; Turner, Adam C

2015-10-01

The use of Monte Carlo simulations in diagnostic medical imaging research is widespread due to its flexibility and ability to estimate quantities that are challenging to measure empirically. However, any new Monte Carlo simulation code needs to be validated before it can be used reliably. The type and degree of validation required depends on the goals of the research project, but, typically, such validation involves either comparison of simulation results to physical measurements or to previously published results obtained with established Monte Carlo codes. The former is complicated due to nuances of experimental conditions and uncertainty, while the latter is challenging due to typical graphical presentation and lack of simulation details in previous publications. In addition, entering the field of Monte Carlo simulations in general involves a steep learning curve. It is not a simple task to learn how to program and interpret a Monte Carlo simulation, even when using one of the publicly available code packages. This Task Group report provides a common reference for benchmarking Monte Carlo simulations across a range of Monte Carlo codes and simulation scenarios. In the report, all simulation conditions are provided for six different Monte Carlo simulation cases that involve common x-ray based imaging research areas. The results obtained for the six cases using four publicly available Monte Carlo software packages are included in tabular form. In addition to a full description of all simulation conditions and results, a discussion and comparison of results among the Monte Carlo packages and the lessons learned during the compilation of these results are included. This abridged version of the report includes only an introductory description of the six cases and a brief example of the results of one of the cases. This work provides an investigator the necessary information to benchmark his/her Monte Carlo simulation software against the reference cases included here before performing his/her own novel research. In addition, an investigator entering the field of Monte Carlo simulations can use these descriptions and results as a self-teaching tool to ensure that he/she is able to perform a specific simulation correctly. Finally, educators can assign these cases as learning projects as part of course objectives or training programs.

Monte Carlo Analysis of Reservoir Models Using Seismic Data and Geostatistical Models

NASA Astrophysics Data System (ADS)

Zunino, A.; Mosegaard, K.; Lange, K.; Melnikova, Y.; Hansen, T. M.

2013-12-01

We present a study on the analysis of petroleum reservoir models consistent with seismic data and geostatistical constraints performed on a synthetic reservoir model. Our aim is to invert directly for structure and rock bulk properties of the target reservoir zone. To infer the rock facies, porosity and oil saturation seismology alone is not sufficient but a rock physics model must be taken into account, which links the unknown properties to the elastic parameters. We then combine a rock physics model with a simple convolutional approach for seismic waves to invert the "measured" seismograms. To solve this inverse problem, we employ a Markov chain Monte Carlo (MCMC) method, because it offers the possibility to handle non-linearity, complex and multi-step forward models and provides realistic estimates of uncertainties. However, for large data sets the MCMC method may be impractical because of a very high computational demand. To face this challenge one strategy is to feed the algorithm with realistic models, hence relying on proper prior information. To address this problem, we utilize an algorithm drawn from geostatistics to generate geologically plausible models which represent samples of the prior distribution. The geostatistical algorithm learns the multiple-point statistics from prototype models (in the form of training images), then generates thousands of different models which are accepted or rejected by a Metropolis sampler. To further reduce the computation time we parallelize the software and run it on multi-core machines. The solution of the inverse problem is then represented by a collection of reservoir models in terms of facies, porosity and oil saturation, which constitute samples of the posterior distribution. We are finally able to produce probability maps of the properties we are interested in by performing statistical analysis on the collection of solutions.

A Probabilistic Framework for Constructing Temporal Relations in Replica Exchange Molecular Trajectories.

PubMed

Chattopadhyay, Aditya; Zheng, Min; Waller, Mark Paul; Priyakumar, U Deva

2018-05-23

Knowledge of the structure and dynamics of biomolecules is essential for elucidating the underlying mechanisms of biological processes. Given the stochastic nature of many biological processes, like protein unfolding, it's almost impossible that two independent simulations will generate the exact same sequence of events, which makes direct analysis of simulations difficult. Statistical models like Markov Chains, transition networks etc. help in shedding some light on the mechanistic nature of such processes by predicting long-time dynamics of these systems from short simulations. However, such methods fall short in analyzing trajectories with partial or no temporal information, for example, replica exchange molecular dynamics or Monte Carlo simulations. In this work we propose a probabilistic algorithm, borrowing concepts from graph theory and machine learning, to extract reactive pathways from molecular trajectories in the absence of temporal data. A suitable vector representation was chosen to represent each frame in the macromolecular trajectory (as a series of interaction and conformational energies) and dimensionality reduction was performed using principal component analysis (PCA). The trajectory was then clustered using a density-based clustering algorithm, where each cluster represents a metastable state on the potential energy surface (PES) of the biomolecule under study. A graph was created with these clusters as nodes with the edges learnt using an iterative expectation maximization algorithm. The most reactive path is conceived as the widest path along this graph. We have tested our method on RNA hairpin unfolding trajectory in aqueous urea solution. Our method makes the understanding of the mechanism of unfolding in RNA hairpin molecule more tractable. As this method doesn't rely on temporal data it can be used to analyze trajectories from Monte Carlo sampling techniques and replica exchange molecular dynamics (REMD).

Bayesian Estimation of Multidimensional Item Response Models. A Comparison of Analytic and Simulation Algorithms

ERIC Educational Resources Information Center

Martin-Fernandez, Manuel; Revuelta, Javier

2017-01-01

This study compares the performance of two estimation algorithms of new usage, the Metropolis-Hastings Robins-Monro (MHRM) and the Hamiltonian MCMC (HMC), with two consolidated algorithms in the psychometric literature, the marginal likelihood via EM algorithm (MML-EM) and the Markov chain Monte Carlo (MCMC), in the estimation of multidimensional…

Fungible Correlation Matrices: A Method for Generating Nonsingular, Singular, and Improper Correlation Matrices for Monte Carlo Research.

PubMed

Waller, Niels G

2016-01-01

For a fixed set of standardized regression coefficients and a fixed coefficient of determination (R-squared), an infinite number of predictor correlation matrices will satisfy the implied quadratic form. I call such matrices fungible correlation matrices. In this article, I describe an algorithm for generating positive definite (PD), positive semidefinite (PSD), or indefinite (ID) fungible correlation matrices that have a random or fixed smallest eigenvalue. The underlying equations of this algorithm are reviewed from both algebraic and geometric perspectives. Two simulation studies illustrate that fungible correlation matrices can be profitably used in Monte Carlo research. The first study uses PD fungible correlation matrices to compare penalized regression algorithms. The second study uses ID fungible correlation matrices to compare matrix-smoothing algorithms. R code for generating fungible correlation matrices is presented in the supplemental materials.

Using Stan for Item Response Theory Models

ERIC Educational Resources Information Center

Ames, Allison J.; Au, Chi Hang

2018-01-01

Stan is a flexible probabilistic programming language providing full Bayesian inference through Hamiltonian Monte Carlo algorithms. The benefits of Hamiltonian Monte Carlo include improved efficiency and faster inference, when compared to other MCMC software implementations. Users can interface with Stan through a variety of computing…

Quantum Monte Carlo Simulation of Frustrated Kondo Lattice Models

NASA Astrophysics Data System (ADS)

Sato, Toshihiro; Assaad, Fakher F.; Grover, Tarun

2018-03-01

The absence of the negative sign problem in quantum Monte Carlo simulations of spin and fermion systems has different origins. World-line based algorithms for spins require positivity of matrix elements whereas auxiliary field approaches for fermions depend on symmetries such as particle-hole symmetry. For negative-sign-free spin and fermionic systems, we show that one can formulate a negative-sign-free auxiliary field quantum Monte Carlo algorithm that allows Kondo coupling of fermions with the spins. Using this general approach, we study a half-filled Kondo lattice model on the honeycomb lattice with geometric frustration. In addition to the conventional Kondo insulator and antiferromagnetically ordered phases, we find a partial Kondo screened state where spins are selectively screened so as to alleviate frustration, and the lattice rotation symmetry is broken nematically.

Parameters Free Computational Characterization of Defects in Transition Metal Oxides with Diffusion Quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Santana, Juan A.; Krogel, Jaron T.; Kent, Paul R.; Reboredo, Fernando

Materials based on transition metal oxides (TMO's) are among the most challenging systems for computational characterization. Reliable and practical computations are possible by directly solving the many-body problem for TMO's with quantum Monte Carlo (QMC) methods. These methods are very computationally intensive, but recent developments in algorithms and computational infrastructures have enabled their application to real materials. We will show our efforts on the application of the diffusion quantum Monte Carlo (DMC) method to study the formation of defects in binary and ternary TMO and heterostructures of TMO. We will also outline current limitations in hardware and algorithms. This work is supported by the Materials Sciences & Engineering Division of the Office of Basic Energy Sciences, U.S. Department of Energy (DOE).

Direct simulation Monte Carlo method for the Uehling-Uhlenbeck-Boltzmann equation.

PubMed

Garcia, Alejandro L; Wagner, Wolfgang

2003-11-01

In this paper we describe a direct simulation Monte Carlo algorithm for the Uehling-Uhlenbeck-Boltzmann equation in terms of Markov processes. This provides a unifying framework for both the classical Boltzmann case as well as the Fermi-Dirac and Bose-Einstein cases. We establish the foundation of the algorithm by demonstrating its link to the kinetic equation. By numerical experiments we study its sensitivity to the number of simulation particles and to the discretization of the velocity space, when approximating the steady-state distribution.

A highly optimized vectorized code for Monte Carlo simulations of SU(3) lattice gauge theories

NASA Technical Reports Server (NTRS)

Barkai, D.; Moriarty, K. J. M.; Rebbi, C.

1984-01-01

New methods are introduced for improving the performance of the vectorized Monte Carlo SU(3) lattice gauge theory algorithm using the CDC CYBER 205. Structure, algorithm and programming considerations are discussed. The performance achieved for a 16(4) lattice on a 2-pipe system may be phrased in terms of the link update time or overall MFLOPS rates. For 32-bit arithmetic, it is 36.3 microsecond/link for 8 hits per iteration (40.9 microsecond for 10 hits) or 101.5 MFLOPS.

Nuclear Forensics Analysis with Missing and Uncertain Data

DOE PAGES

Langan, Roisin T.; Archibald, Richard K.; Lamberti, Vincent

2015-10-05

We have applied a new imputation-based method for analyzing incomplete data, called Monte Carlo Bayesian Database Generation (MCBDG), to the Spent Fuel Isotopic Composition (SFCOMPO) database. About 60% of the entries are absent for SFCOMPO. The method estimates missing values of a property from a probability distribution created from the existing data for the property, and then generates multiple instances of the completed database for training a machine learning algorithm. Uncertainty in the data is represented by an empirical or an assumed error distribution. The method makes few assumptions about the underlying data, and compares favorably against results obtained bymore » replacing missing information with constant values.« less

Fullrmc, a rigid body Reverse Monte Carlo modeling package enabled with machine learning and artificial intelligence.

PubMed

Aoun, Bachir

2016-05-05

A new Reverse Monte Carlo (RMC) package "fullrmc" for atomic or rigid body and molecular, amorphous, or crystalline materials is presented. fullrmc main purpose is to provide a fully modular, fast and flexible software, thoroughly documented, complex molecules enabled, written in a modern programming language (python, cython, C and C++ when performance is needed) and complying to modern programming practices. fullrmc approach in solving an atomic or molecular structure is different from existing RMC algorithms and software. In a nutshell, traditional RMC methods and software randomly adjust atom positions until the whole system has the greatest consistency with a set of experimental data. In contrast, fullrmc applies smart moves endorsed with reinforcement machine learning to groups of atoms. While fullrmc allows running traditional RMC modeling, the uniqueness of this approach resides in its ability to customize grouping atoms in any convenient way with no additional programming efforts and to apply smart and more physically meaningful moves to the defined groups of atoms. In addition, fullrmc provides a unique way with almost no additional computational cost to recur a group's selection, allowing the system to go out of local minimas by refining a group's position or exploring through and beyond not allowed positions and energy barriers the unrestricted three dimensional space around a group. © 2016 Wiley Periodicals, Inc.

Fullrmc, a rigid body reverse monte carlo modeling package enabled with machine learning and artificial intelligence

DOE PAGES

Aoun, Bachir

2016-01-22

Here, a new Reverse Monte Carlo (RMC) package ‘fullrmc’ for atomic or rigid body and molecular, amorphous or crystalline materials is presented. fullrmc main purpose is to provide a fully modular, fast and flexible software, thoroughly documented, complex molecules enabled, written in a modern programming language (python, cython ,C and C++ when performance is needed) and complying to modern programming practices. fullrmc approach in solving an atomic or molecular structure is different from existing RMC algorithms and software. In a nutshell, traditional RMC methods and software randomly adjust atom positions until the whole system has the greatest consistency with amore » set of experimental data. In contrast, fullrmc applies smart moves endorsed with reinforcement machine learning to groups of atoms. While fullrmc allows running traditional RMC modelling, the uniqueness of this approach resides in its ability to customize grouping atoms in any convenient way with no additional programming efforts and to apply smart and more physically meaningful moves to the defined groups of atoms. Also fullrmc provides a unique way with almost no additional computational cost to recur a group’s selection, allowing the system to go out of local minimas by refining a group’s position or exploring through and beyond not allowed positions and energy barriers the unrestricted three dimensional space around a group.« less

Fullrmc, a rigid body reverse monte carlo modeling package enabled with machine learning and artificial intelligence

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

Aoun, Bachir

Here, a new Reverse Monte Carlo (RMC) package ‘fullrmc’ for atomic or rigid body and molecular, amorphous or crystalline materials is presented. fullrmc main purpose is to provide a fully modular, fast and flexible software, thoroughly documented, complex molecules enabled, written in a modern programming language (python, cython ,C and C++ when performance is needed) and complying to modern programming practices. fullrmc approach in solving an atomic or molecular structure is different from existing RMC algorithms and software. In a nutshell, traditional RMC methods and software randomly adjust atom positions until the whole system has the greatest consistency with amore » set of experimental data. In contrast, fullrmc applies smart moves endorsed with reinforcement machine learning to groups of atoms. While fullrmc allows running traditional RMC modelling, the uniqueness of this approach resides in its ability to customize grouping atoms in any convenient way with no additional programming efforts and to apply smart and more physically meaningful moves to the defined groups of atoms. Also fullrmc provides a unique way with almost no additional computational cost to recur a group’s selection, allowing the system to go out of local minimas by refining a group’s position or exploring through and beyond not allowed positions and energy barriers the unrestricted three dimensional space around a group.« less

Delayed Slater determinant update algorithms for high efficiency quantum Monte Carlo

DOE PAGES

McDaniel, Tyler; D’Azevedo, Ed F.; Li, Ying Wai; ...

2017-11-07

Within ab initio Quantum Monte Carlo simulations, the leading numerical cost for large systems is the computation of the values of the Slater determinants in the trial wavefunction. Each Monte Carlo step requires finding the determinant of a dense matrix. This is most commonly iteratively evaluated using a rank-1 Sherman-Morrison updating scheme to avoid repeated explicit calculation of the inverse. The overall computational cost is therefore formally cubic in the number of electrons or matrix size. To improve the numerical efficiency of this procedure, we propose a novel multiple rank delayed update scheme. This strategy enables probability evaluation with applicationmore » of accepted moves to the matrices delayed until after a predetermined number of moves, K. The accepted events are then applied to the matrices en bloc with enhanced arithmetic intensity and computational efficiency via matrix-matrix operations instead of matrix-vector operations. Here this procedure does not change the underlying Monte Carlo sampling or its statistical efficiency. For calculations on large systems and algorithms such as diffusion Monte Carlo where the acceptance ratio is high, order of magnitude improvements in the update time can be obtained on both multi- core CPUs and GPUs.« less

Delayed Slater determinant update algorithms for high efficiency quantum Monte Carlo

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

McDaniel, Tyler; D’Azevedo, Ed F.; Li, Ying Wai

Within ab initio Quantum Monte Carlo simulations, the leading numerical cost for large systems is the computation of the values of the Slater determinants in the trial wavefunction. Each Monte Carlo step requires finding the determinant of a dense matrix. This is most commonly iteratively evaluated using a rank-1 Sherman-Morrison updating scheme to avoid repeated explicit calculation of the inverse. The overall computational cost is therefore formally cubic in the number of electrons or matrix size. To improve the numerical efficiency of this procedure, we propose a novel multiple rank delayed update scheme. This strategy enables probability evaluation with applicationmore » of accepted moves to the matrices delayed until after a predetermined number of moves, K. The accepted events are then applied to the matrices en bloc with enhanced arithmetic intensity and computational efficiency via matrix-matrix operations instead of matrix-vector operations. Here this procedure does not change the underlying Monte Carlo sampling or its statistical efficiency. For calculations on large systems and algorithms such as diffusion Monte Carlo where the acceptance ratio is high, order of magnitude improvements in the update time can be obtained on both multi- core CPUs and GPUs.« less

Delayed Slater determinant update algorithms for high efficiency quantum Monte Carlo.

PubMed

McDaniel, T; D'Azevedo, E F; Li, Y W; Wong, K; Kent, P R C

2017-11-07

Within ab initio Quantum Monte Carlo simulations, the leading numerical cost for large systems is the computation of the values of the Slater determinants in the trial wavefunction. Each Monte Carlo step requires finding the determinant of a dense matrix. This is most commonly iteratively evaluated using a rank-1 Sherman-Morrison updating scheme to avoid repeated explicit calculation of the inverse. The overall computational cost is, therefore, formally cubic in the number of electrons or matrix size. To improve the numerical efficiency of this procedure, we propose a novel multiple rank delayed update scheme. This strategy enables probability evaluation with an application of accepted moves to the matrices delayed until after a predetermined number of moves, K. The accepted events are then applied to the matrices en bloc with enhanced arithmetic intensity and computational efficiency via matrix-matrix operations instead of matrix-vector operations. This procedure does not change the underlying Monte Carlo sampling or its statistical efficiency. For calculations on large systems and algorithms such as diffusion Monte Carlo, where the acceptance ratio is high, order of magnitude improvements in the update time can be obtained on both multi-core central processing units and graphical processing units.

Delayed Slater determinant update algorithms for high efficiency quantum Monte Carlo

NASA Astrophysics Data System (ADS)

McDaniel, T.; D'Azevedo, E. F.; Li, Y. W.; Wong, K.; Kent, P. R. C.

2017-11-01

Within ab initio Quantum Monte Carlo simulations, the leading numerical cost for large systems is the computation of the values of the Slater determinants in the trial wavefunction. Each Monte Carlo step requires finding the determinant of a dense matrix. This is most commonly iteratively evaluated using a rank-1 Sherman-Morrison updating scheme to avoid repeated explicit calculation of the inverse. The overall computational cost is, therefore, formally cubic in the number of electrons or matrix size. To improve the numerical efficiency of this procedure, we propose a novel multiple rank delayed update scheme. This strategy enables probability evaluation with an application of accepted moves to the matrices delayed until after a predetermined number of moves, K. The accepted events are then applied to the matrices en bloc with enhanced arithmetic intensity and computational efficiency via matrix-matrix operations instead of matrix-vector operations. This procedure does not change the underlying Monte Carlo sampling or its statistical efficiency. For calculations on large systems and algorithms such as diffusion Monte Carlo, where the acceptance ratio is high, order of magnitude improvements in the update time can be obtained on both multi-core central processing units and graphical processing units.

Distance between configurations in Markov chain Monte Carlo simulations

NASA Astrophysics Data System (ADS)

Fukuma, Masafumi; Matsumoto, Nobuyuki; Umeda, Naoya

2017-12-01

For a given Markov chain Monte Carlo algorithm we introduce a distance between two configurations that quantifies the difficulty of transition from one configuration to the other configuration. We argue that the distance takes a universal form for the class of algorithms which generate local moves in the configuration space. We explicitly calculate the distance for the Langevin algorithm, and show that it certainly has desired and expected properties as distance. We further show that the distance for a multimodal distribution gets dramatically reduced from a large value by the introduction of a tempering method. We also argue that, when the original distribution is highly multimodal with large number of degenerate vacua, an anti-de Sitter-like geometry naturally emerges in the extended configuration space.

The impact of low-Z and high-Z metal implants in IMRT: A Monte Carlo study of dose inaccuracies in commercial dose algorithms

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

Spadea, Maria Francesca, E-mail: mfspadea@unicz.it; Verburg, Joost Mathias; Seco, Joao

2014-01-15

Purpose: The aim of the study was to evaluate the dosimetric impact of low-Z and high-Z metallic implants on IMRT plans. Methods: Computed tomography (CT) scans of three patients were analyzed to study effects due to the presence of Titanium (low-Z), Platinum and Gold (high-Z) inserts. To eliminate artifacts in CT images, a sinogram-based metal artifact reduction algorithm was applied. IMRT dose calculations were performed on both the uncorrected and corrected images using a commercial planning system (convolution/superposition algorithm) and an in-house Monte Carlo platform. Dose differences between uncorrected and corrected datasets were computed and analyzed using gamma index (Pγ{submore » <1}) and setting 2 mm and 2% as distance to agreement and dose difference criteria, respectively. Beam specific depth dose profiles across the metal were also examined. Results: Dose discrepancies between corrected and uncorrected datasets were not significant for low-Z material. High-Z materials caused under-dosage of 20%–25% in the region surrounding the metal and over dosage of 10%–15% downstream of the hardware. Gamma index test yielded Pγ{sub <1}>99% for all low-Z cases; while for high-Z cases it returned 91% < Pγ{sub <1}< 99%. Analysis of the depth dose curve of a single beam for low-Z cases revealed that, although the dose attenuation is altered inside the metal, it does not differ downstream of the insert. However, for high-Z metal implants the dose is increased up to 10%–12% around the insert. In addition, Monte Carlo method was more sensitive to the presence of metal inserts than superposition/convolution algorithm. Conclusions: The reduction in terms of dose of metal artifacts in CT images is relevant for high-Z implants. In this case, dose distribution should be calculated using Monte Carlo algorithms, given their superior accuracy in dose modeling in and around the metal. In addition, the knowledge of the composition of metal inserts improves the accuracy of the Monte Carlo dose calculation significantly.« less

A resilient and efficient CFD framework: Statistical learning tools for multi-fidelity and heterogeneous information fusion

NASA Astrophysics Data System (ADS)

Lee, Seungjoon; Kevrekidis, Ioannis G.; Karniadakis, George Em

2017-09-01

Exascale-level simulations require fault-resilient algorithms that are robust against repeated and expected software and/or hardware failures during computations, which may render the simulation results unsatisfactory. If each processor can share some global information about the simulation from a coarse, limited accuracy but relatively costless auxiliary simulator we can effectively fill-in the missing spatial data at the required times by a statistical learning technique - multi-level Gaussian process regression, on the fly; this has been demonstrated in previous work [1]. Based on the previous work, we also employ another (nonlinear) statistical learning technique, Diffusion Maps, that detects computational redundancy in time and hence accelerate the simulation by projective time integration, giving the overall computation a "patch dynamics" flavor. Furthermore, we are now able to perform information fusion with multi-fidelity and heterogeneous data (including stochastic data). Finally, we set the foundations of a new framework in CFD, called patch simulation, that combines information fusion techniques from, in principle, multiple fidelity and resolution simulations (and even experiments) with a new adaptive timestep refinement technique. We present two benchmark problems (the heat equation and the Navier-Stokes equations) to demonstrate the new capability that statistical learning tools can bring to traditional scientific computing algorithms. For each problem, we rely on heterogeneous and multi-fidelity data, either from a coarse simulation of the same equation or from a stochastic, particle-based, more "microscopic" simulation. We consider, as such "auxiliary" models, a Monte Carlo random walk for the heat equation and a dissipative particle dynamics (DPD) model for the Navier-Stokes equations. More broadly, in this paper we demonstrate the symbiotic and synergistic combination of statistical learning, domain decomposition, and scientific computing in exascale simulations.

Moral foundations in an interacting neural networks society: A statistical mechanics analysis

NASA Astrophysics Data System (ADS)

Vicente, R.; Susemihl, A.; Jericó, J. P.; Caticha, N.

2014-04-01

The moral foundations theory supports that people, across cultures, tend to consider a small number of dimensions when classifying issues on a moral basis. The data also show that the statistics of weights attributed to each moral dimension is related to self-declared political affiliation, which in turn has been connected to cognitive learning styles by the recent literature in neuroscience and psychology. Inspired by these data, we propose a simple statistical mechanics model with interacting neural networks classifying vectors and learning from members of their social neighbourhood about their average opinion on a large set of issues. The purpose of learning is to reduce dissension among agents when disagreeing. We consider a family of learning algorithms parametrized by δ, that represents the importance given to corroborating (same sign) opinions. We define an order parameter that quantifies the diversity of opinions in a group with homogeneous learning style. Using Monte Carlo simulations and a mean field approximation we find the relation between the order parameter and the learning parameter δ at a temperature we associate with the importance of social influence in a given group. In concordance with data, groups that rely more strongly on corroborating evidence sustain less opinion diversity. We discuss predictions of the model and propose possible experimental tests.

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

Jakeman, John D.; Narayan, Akil; Zhou, Tao

We propose an algorithm for recovering sparse orthogonal polynomial expansions via collocation. A standard sampling approach for recovering sparse polynomials uses Monte Carlo sampling, from the density of orthogonality, which results in poor function recovery when the polynomial degree is high. Our proposed approach aims to mitigate this limitation by sampling with respect to the weighted equilibrium measure of the parametric domain and subsequently solves a preconditionedmore » $$\\ell^1$$-minimization problem, where the weights of the diagonal preconditioning matrix are given by evaluations of the Christoffel function. Our algorithm can be applied to a wide class of orthogonal polynomial families on bounded and unbounded domains, including all classical families. We present theoretical analysis to motivate the algorithm and numerical results that show our method is superior to standard Monte Carlo methods in many situations of interest. In conclusion, numerical examples are also provided to demonstrate that our proposed algorithm leads to comparable or improved accuracy even when compared with Legendre- and Hermite-specific algorithms.« less

Hybrid-optimization strategy for the communication of large-scale Kinetic Monte Carlo simulation

NASA Astrophysics Data System (ADS)

Wu, Baodong; Li, Shigang; Zhang, Yunquan; Nie, Ningming

2017-02-01

The parallel Kinetic Monte Carlo (KMC) algorithm based on domain decomposition has been widely used in large-scale physical simulations. However, the communication overhead of the parallel KMC algorithm is critical, and severely degrades the overall performance and scalability. In this paper, we present a hybrid optimization strategy to reduce the communication overhead for the parallel KMC simulations. We first propose a communication aggregation algorithm to reduce the total number of messages and eliminate the communication redundancy. Then, we utilize the shared memory to reduce the memory copy overhead of the intra-node communication. Finally, we optimize the communication scheduling using the neighborhood collective operations. We demonstrate the scalability and high performance of our hybrid optimization strategy by both theoretical and experimental analysis. Results show that the optimized KMC algorithm exhibits better performance and scalability than the well-known open-source library-SPPARKS. On 32-node Xeon E5-2680 cluster (total 640 cores), the optimized algorithm reduces the communication time by 24.8% compared with SPPARKS.

A Generalized Sampling and Preconditioning Scheme for Sparse Approximation of Polynomial Chaos Expansions

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

Jakeman, John D.; Narayan, Akil; Zhou, Tao

We propose an algorithm for recovering sparse orthogonal polynomial expansions via collocation. A standard sampling approach for recovering sparse polynomials uses Monte Carlo sampling, from the density of orthogonality, which results in poor function recovery when the polynomial degree is high. Our proposed approach aims to mitigate this limitation by sampling with respect to the weighted equilibrium measure of the parametric domain and subsequently solves a preconditionedmore » $$\\ell^1$$-minimization problem, where the weights of the diagonal preconditioning matrix are given by evaluations of the Christoffel function. Our algorithm can be applied to a wide class of orthogonal polynomial families on bounded and unbounded domains, including all classical families. We present theoretical analysis to motivate the algorithm and numerical results that show our method is superior to standard Monte Carlo methods in many situations of interest. In conclusion, numerical examples are also provided to demonstrate that our proposed algorithm leads to comparable or improved accuracy even when compared with Legendre- and Hermite-specific algorithms.« less

A Generalized Sampling and Preconditioning Scheme for Sparse Approximation of Polynomial Chaos Expansions

DOE PAGES

Jakeman, John D.; Narayan, Akil; Zhou, Tao

2017-06-22

We propose an algorithm for recovering sparse orthogonal polynomial expansions via collocation. A standard sampling approach for recovering sparse polynomials uses Monte Carlo sampling, from the density of orthogonality, which results in poor function recovery when the polynomial degree is high. Our proposed approach aims to mitigate this limitation by sampling with respect to the weighted equilibrium measure of the parametric domain and subsequently solves a preconditionedmore » $$\\ell^1$$-minimization problem, where the weights of the diagonal preconditioning matrix are given by evaluations of the Christoffel function. Our algorithm can be applied to a wide class of orthogonal polynomial families on bounded and unbounded domains, including all classical families. We present theoretical analysis to motivate the algorithm and numerical results that show our method is superior to standard Monte Carlo methods in many situations of interest. In conclusion, numerical examples are also provided to demonstrate that our proposed algorithm leads to comparable or improved accuracy even when compared with Legendre- and Hermite-specific algorithms.« less

Sediment classification using neural networks: An example from the site-U1344A of IODP Expedition 323 in the Bering Sea

NASA Astrophysics Data System (ADS)

Ojha, Maheswar; Maiti, Saumen

2016-03-01

A novel approach based on the concept of Bayesian neural network (BNN) has been implemented for classifying sediment boundaries using downhole log data obtained during Integrated Ocean Drilling Program (IODP) Expedition 323 in the Bering Sea slope region. The Bayesian framework in conjunction with Markov Chain Monte Carlo (MCMC)/hybrid Monte Carlo (HMC) learning paradigm has been applied to constrain the lithology boundaries using density, density porosity, gamma ray, sonic P-wave velocity and electrical resistivity at the Hole U1344A. We have demonstrated the effectiveness of our supervised classification methodology by comparing our findings with a conventional neural network and a Bayesian neural network optimized by scaled conjugate gradient method (SCG), and tested the robustness of the algorithm in the presence of red noise in the data. The Bayesian results based on the HMC algorithm (BNN.HMC) resolve detailed finer structures at certain depths in addition to main lithology such as silty clay, diatom clayey silt and sandy silt. Our method also recovers the lithology information from a depth ranging between 615 and 655 m Wireline log Matched depth below Sea Floor of no core recovery zone. Our analyses demonstrate that the BNN based approach renders robust means for the classification of complex lithology successions at the Hole U1344A, which could be very useful for other studies and understanding the oceanic crustal inhomogeneity and structural discontinuities.

An Ensemble-Based Smoother with Retrospectively Updated Weights for Highly Nonlinear Systems

NASA Technical Reports Server (NTRS)

Chin, T. M.; Turmon, M. J.; Jewell, J. B.; Ghil, M.

2006-01-01

Monte Carlo computational methods have been introduced into data assimilation for nonlinear systems in order to alleviate the computational burden of updating and propagating the full probability distribution. By propagating an ensemble of representative states, algorithms like the ensemble Kalman filter (EnKF) and the resampled particle filter (RPF) rely on the existing modeling infrastructure to approximate the distribution based on the evolution of this ensemble. This work presents an ensemble-based smoother that is applicable to the Monte Carlo filtering schemes like EnKF and RPF. At the minor cost of retrospectively updating a set of weights for ensemble members, this smoother has demonstrated superior capabilities in state tracking for two highly nonlinear problems: the double-well potential and trivariate Lorenz systems. The algorithm does not require retrospective adaptation of the ensemble members themselves, and it is thus suited to a streaming operational mode. The accuracy of the proposed backward-update scheme in estimating non-Gaussian distributions is evaluated by comparison to the more accurate estimates provided by a Markov chain Monte Carlo algorithm.

Particle Communication and Domain Neighbor Coupling: Scalable Domain Decomposed Algorithms for Monte Carlo Particle Transport

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

O'Brien, M. J.; Brantley, P. S.

2015-01-20

In order to run Monte Carlo particle transport calculations on new supercomputers with hundreds of thousands or millions of processors, care must be taken to implement scalable algorithms. This means that the algorithms must continue to perform well as the processor count increases. In this paper, we examine the scalability of:(1) globally resolving the particle locations on the correct processor, (2) deciding that particle streaming communication has finished, and (3) efficiently coupling neighbor domains together with different replication levels. We have run domain decomposed Monte Carlo particle transport on up to 2 21 = 2,097,152 MPI processes on the IBMmore » BG/Q Sequoia supercomputer and observed scalable results that agree with our theoretical predictions. These calculations were carefully constructed to have the same amount of work on every processor, i.e. the calculation is already load balanced. We also examine load imbalanced calculations where each domain’s replication level is proportional to its particle workload. In this case we show how to efficiently couple together adjacent domains to maintain within workgroup load balance and minimize memory usage.« less

Development of reversible jump Markov Chain Monte Carlo algorithm in the Bayesian mixture modeling for microarray data in Indonesia

NASA Astrophysics Data System (ADS)

Astuti, Ani Budi; Iriawan, Nur; Irhamah, Kuswanto, Heri

2017-12-01

In the Bayesian mixture modeling requires stages the identification number of the most appropriate mixture components thus obtained mixture models fit the data through data driven concept. Reversible Jump Markov Chain Monte Carlo (RJMCMC) is a combination of the reversible jump (RJ) concept and the Markov Chain Monte Carlo (MCMC) concept used by some researchers to solve the problem of identifying the number of mixture components which are not known with certainty number. In its application, RJMCMC using the concept of the birth/death and the split-merge with six types of movement, that are w updating, θ updating, z updating, hyperparameter β updating, split-merge for components and birth/death from blank components. The development of the RJMCMC algorithm needs to be done according to the observed case. The purpose of this study is to know the performance of RJMCMC algorithm development in identifying the number of mixture components which are not known with certainty number in the Bayesian mixture modeling for microarray data in Indonesia. The results of this study represent that the concept RJMCMC algorithm development able to properly identify the number of mixture components in the Bayesian normal mixture model wherein the component mixture in the case of microarray data in Indonesia is not known for certain number.

Modelling maximum river flow by using Bayesian Markov Chain Monte Carlo

NASA Astrophysics Data System (ADS)

Cheong, R. Y.; Gabda, D.

2017-09-01

Analysis of flood trends is vital since flooding threatens human living in terms of financial, environment and security. The data of annual maximum river flows in Sabah were fitted into generalized extreme value (GEV) distribution. Maximum likelihood estimator (MLE) raised naturally when working with GEV distribution. However, previous researches showed that MLE provide unstable results especially in small sample size. In this study, we used different Bayesian Markov Chain Monte Carlo (MCMC) based on Metropolis-Hastings algorithm to estimate GEV parameters. Bayesian MCMC method is a statistical inference which studies the parameter estimation by using posterior distribution based on Bayes’ theorem. Metropolis-Hastings algorithm is used to overcome the high dimensional state space faced in Monte Carlo method. This approach also considers more uncertainty in parameter estimation which then presents a better prediction on maximum river flow in Sabah.

Game playing.

PubMed

Rosin, Christopher D

2014-03-01

Game playing has been a core domain of artificial intelligence research since the beginnings of the field. Game playing provides clearly defined arenas within which computational approaches can be readily compared to human expertise through head-to-head competition and other benchmarks. Game playing research has identified several simple core algorithms that provide successful foundations, with development focused on the challenges of defeating human experts in specific games. Key developments include minimax search in chess, machine learning from self-play in backgammon, and Monte Carlo tree search in Go. These approaches have generalized successfully to additional games. While computers have surpassed human expertise in a wide variety of games, open challenges remain and research focuses on identifying and developing new successful algorithmic foundations. WIREs Cogn Sci 2014, 5:193-205. doi: 10.1002/wcs.1278 CONFLICT OF INTEREST: The author has declared no conflicts of interest for this article. For further resources related to this article, please visit the WIREs website. © 2014 John Wiley & Sons, Ltd.

Simulated Stochastic Approximation Annealing for Global Optimization with a Square-Root Cooling Schedule

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

Liang, Faming; Cheng, Yichen; Lin, Guang

2014-06-13

Simulated annealing has been widely used in the solution of optimization problems. As known by many researchers, the global optima cannot be guaranteed to be located by simulated annealing unless a logarithmic cooling schedule is used. However, the logarithmic cooling schedule is so slow that no one can afford to have such a long CPU time. This paper proposes a new stochastic optimization algorithm, the so-called simulated stochastic approximation annealing algorithm, which is a combination of simulated annealing and the stochastic approximation Monte Carlo algorithm. Under the framework of stochastic approximation Markov chain Monte Carlo, it is shown that themore » new algorithm can work with a cooling schedule in which the temperature can decrease much faster than in the logarithmic cooling schedule, e.g., a square-root cooling schedule, while guaranteeing the global optima to be reached when the temperature tends to zero. The new algorithm has been tested on a few benchmark optimization problems, including feed-forward neural network training and protein-folding. The numerical results indicate that the new algorithm can significantly outperform simulated annealing and other competitors.« less

Discrete Diffusion Monte Carlo for Electron Thermal Transport

NASA Astrophysics Data System (ADS)

Chenhall, Jeffrey; Cao, Duc; Wollaeger, Ryan; Moses, Gregory

2014-10-01

The iSNB (implicit Schurtz Nicolai Busquet electron thermal transport method of Cao et al. is adapted to a Discrete Diffusion Monte Carlo (DDMC) solution method for eventual inclusion in a hybrid IMC-DDMC (Implicit Monte Carlo) method. The hybrid method will combine the efficiency of a diffusion method in short mean free path regions with the accuracy of a transport method in long mean free path regions. The Monte Carlo nature of the approach allows the algorithm to be massively parallelized. Work to date on the iSNB-DDMC method will be presented. This work was supported by Sandia National Laboratory - Albuquerque.

Monte Carlo treatment planning with modulated electron radiotherapy: framework development and application

NASA Astrophysics Data System (ADS)

Alexander, Andrew William

Within the field of medical physics, Monte Carlo radiation transport simulations are considered to be the most accurate method for the determination of dose distributions in patients. The McGill Monte Carlo treatment planning system (MMCTP), provides a flexible software environment to integrate Monte Carlo simulations with current and new treatment modalities. A developing treatment modality called energy and intensity modulated electron radiotherapy (MERT) is a promising modality, which has the fundamental capabilities to enhance the dosimetry of superficial targets. An objective of this work is to advance the research and development of MERT with the end goal of clinical use. To this end, we present the MMCTP system with an integrated toolkit for MERT planning and delivery of MERT fields. Delivery is achieved using an automated "few leaf electron collimator" (FLEC) and a controller. Aside from the MERT planning toolkit, the MMCTP system required numerous add-ons to perform the complex task of large-scale autonomous Monte Carlo simulations. The first was a DICOM import filter, followed by the implementation of DOSXYZnrc as a dose calculation engine and by logic methods for submitting and updating the status of Monte Carlo simulations. Within this work we validated the MMCTP system with a head and neck Monte Carlo recalculation study performed by a medical dosimetrist. The impact of MMCTP lies in the fact that it allows for systematic and platform independent large-scale Monte Carlo dose calculations for different treatment sites and treatment modalities. In addition to the MERT planning tools, various optimization algorithms were created external to MMCTP. The algorithms produced MERT treatment plans based on dose volume constraints that employ Monte Carlo pre-generated patient-specific kernels. The Monte Carlo kernels are generated from patient-specific Monte Carlo dose distributions within MMCTP. The structure of the MERT planning toolkit software and optimization algorithms are demonstrated. We investigated the clinical significance of MERT on spinal irradiation, breast boost irradiation, and a head and neck sarcoma cancer site using several parameters to analyze the treatment plans. Finally, we investigated the idea of mixed beam photon and electron treatment planning. Photon optimization treatment planning tools were included within the MERT planning toolkit for the purpose of mixed beam optimization. In conclusion, this thesis work has resulted in the development of an advanced framework for photon and electron Monte Carlo treatment planning studies and the development of an inverse planning system for photon, electron or mixed beam radiotherapy (MBRT). The justification and validation of this work is found within the results of the planning studies, which have demonstrated dosimetric advantages to using MERT or MBRT in comparison to clinical treatment alternatives.

Predicting Protein-protein Association Rates using Coarse-grained Simulation and Machine Learning

NASA Astrophysics Data System (ADS)

Xie, Zhong-Ru; Chen, Jiawen; Wu, Yinghao

2017-04-01

Protein-protein interactions dominate all major biological processes in living cells. We have developed a new Monte Carlo-based simulation algorithm to study the kinetic process of protein association. We tested our method on a previously used large benchmark set of 49 protein complexes. The predicted rate was overestimated in the benchmark test compared to the experimental results for a group of protein complexes. We hypothesized that this resulted from molecular flexibility at the interface regions of the interacting proteins. After applying a machine learning algorithm with input variables that accounted for both the conformational flexibility and the energetic factor of binding, we successfully identified most of the protein complexes with overestimated association rates and improved our final prediction by using a cross-validation test. This method was then applied to a new independent test set and resulted in a similar prediction accuracy to that obtained using the training set. It has been thought that diffusion-limited protein association is dominated by long-range interactions. Our results provide strong evidence that the conformational flexibility also plays an important role in regulating protein association. Our studies provide new insights into the mechanism of protein association and offer a computationally efficient tool for predicting its rate.

Predicting Protein–protein Association Rates using Coarse-grained Simulation and Machine Learning

PubMed Central

Xie, Zhong-Ru; Chen, Jiawen; Wu, Yinghao

2017-01-01

Protein–protein interactions dominate all major biological processes in living cells. We have developed a new Monte Carlo-based simulation algorithm to study the kinetic process of protein association. We tested our method on a previously used large benchmark set of 49 protein complexes. The predicted rate was overestimated in the benchmark test compared to the experimental results for a group of protein complexes. We hypothesized that this resulted from molecular flexibility at the interface regions of the interacting proteins. After applying a machine learning algorithm with input variables that accounted for both the conformational flexibility and the energetic factor of binding, we successfully identified most of the protein complexes with overestimated association rates and improved our final prediction by using a cross-validation test. This method was then applied to a new independent test set and resulted in a similar prediction accuracy to that obtained using the training set. It has been thought that diffusion-limited protein association is dominated by long-range interactions. Our results provide strong evidence that the conformational flexibility also plays an important role in regulating protein association. Our studies provide new insights into the mechanism of protein association and offer a computationally efficient tool for predicting its rate. PMID:28418043

Predicting Protein-protein Association Rates using Coarse-grained Simulation and Machine Learning.

PubMed

Xie, Zhong-Ru; Chen, Jiawen; Wu, Yinghao

2017-04-18

Protein-protein interactions dominate all major biological processes in living cells. We have developed a new Monte Carlo-based simulation algorithm to study the kinetic process of protein association. We tested our method on a previously used large benchmark set of 49 protein complexes. The predicted rate was overestimated in the benchmark test compared to the experimental results for a group of protein complexes. We hypothesized that this resulted from molecular flexibility at the interface regions of the interacting proteins. After applying a machine learning algorithm with input variables that accounted for both the conformational flexibility and the energetic factor of binding, we successfully identified most of the protein complexes with overestimated association rates and improved our final prediction by using a cross-validation test. This method was then applied to a new independent test set and resulted in a similar prediction accuracy to that obtained using the training set. It has been thought that diffusion-limited protein association is dominated by long-range interactions. Our results provide strong evidence that the conformational flexibility also plays an important role in regulating protein association. Our studies provide new insights into the mechanism of protein association and offer a computationally efficient tool for predicting its rate.

Simulation-Based Model Checking for Nondeterministic Systems and Rare Events

DTIC Science & Technology

2016-03-24

year, we have investigated AO* search and Monte Carlo Tree Search algorithms to complement and enhance CMU’s SMCMDP. 1 Final Report, March 14... tree , so we can use it to find the probability of reachability for a property in PRISM’s Probabilistic LTL. By finding the maximum probability of...savings, particularly when handling very large models. 2.3 Monte Carlo Tree Search The Monte Carlo sampling process in SMCMDP can take a long time to

WARP

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

Bergmann, Ryan M.; Rowland, Kelly L.

2017-04-12

WARP, which can stand for ``Weaving All the Random Particles,'' is a three-dimensional (3D) continuous energy Monte Carlo neutron transport code developed at UC Berkeley to efficiently execute on NVIDIA graphics processing unit (GPU) platforms. WARP accelerates Monte Carlo simulations while preserving the benefits of using the Monte Carlo method, namely, that very few physical and geometrical simplifications are applied. WARP is able to calculate multiplication factors, neutron flux distributions (in both space and energy), and fission source distributions for time-independent neutron transport problems. It can run in both criticality or fixed source modes, but fixed source mode is currentlymore » not robust, optimized, or maintained in the newest version. WARP can transport neutrons in unrestricted arrangements of parallelepipeds, hexagonal prisms, cylinders, and spheres. The goal of developing WARP is to investigate algorithms that can grow into a full-featured, continuous energy, Monte Carlo neutron transport code that is accelerated by running on GPUs. The crux of the effort is to make Monte Carlo calculations faster while producing accurate results. Modern supercomputers are commonly being built with GPU coprocessor cards in their nodes to increase their computational efficiency and performance. GPUs execute efficiently on data-parallel problems, but most CPU codes, including those for Monte Carlo neutral particle transport, are predominantly task-parallel. WARP uses a data-parallel neutron transport algorithm to take advantage of the computing power GPUs offer.« less

Virtual Network Embedding via Monte Carlo Tree Search.

PubMed

Haeri, Soroush; Trajkovic, Ljiljana

2018-02-01

Network virtualization helps overcome shortcomings of the current Internet architecture. The virtualized network architecture enables coexistence of multiple virtual networks (VNs) on an existing physical infrastructure. VN embedding (VNE) problem, which deals with the embedding of VN components onto a physical network, is known to be -hard. In this paper, we propose two VNE algorithms: MaVEn-M and MaVEn-S. MaVEn-M employs the multicommodity flow algorithm for virtual link mapping while MaVEn-S uses the shortest-path algorithm. They formalize the virtual node mapping problem by using the Markov decision process (MDP) framework and devise action policies (node mappings) for the proposed MDP using the Monte Carlo tree search algorithm. Service providers may adjust the execution time of the MaVEn algorithms based on the traffic load of VN requests. The objective of the algorithms is to maximize the profit of infrastructure providers. We develop a discrete event VNE simulator to implement and evaluate performance of MaVEn-M, MaVEn-S, and several recently proposed VNE algorithms. We introduce profitability as a new performance metric that captures both acceptance and revenue to cost ratios. Simulation results show that the proposed algorithms find more profitable solutions than the existing algorithms. Given additional computation time, they further improve embedding solutions.

Monte Carlo sampling in diffusive dynamical systems

NASA Astrophysics Data System (ADS)

Tapias, Diego; Sanders, David P.; Altmann, Eduardo G.

2018-05-01

We introduce a Monte Carlo algorithm to efficiently compute transport properties of chaotic dynamical systems. Our method exploits the importance sampling technique that favors trajectories in the tail of the distribution of displacements, where deviations from a diffusive process are most prominent. We search for initial conditions using a proposal that correlates states in the Markov chain constructed via a Metropolis-Hastings algorithm. We show that our method outperforms the direct sampling method and also Metropolis-Hastings methods with alternative proposals. We test our general method through numerical simulations in 1D (box-map) and 2D (Lorentz gas) systems.

Vectorization of a Monte Carlo simulation scheme for nonequilibrium gas dynamics

NASA Technical Reports Server (NTRS)

Boyd, Iain D.

1991-01-01

Significant improvement has been obtained in the numerical performance of a Monte Carlo scheme for the analysis of nonequilibrium gas dynamics through an implementation of the algorithm which takes advantage of vector hardware, as presently demonstrated through application to three different problems. These are (1) a 1D standing-shock wave; (2) the flow of an expanding gas through an axisymmetric nozzle; and (3) the hypersonic flow of Ar gas over a 3D wedge. Problem (3) is illustrative of the greatly increased number of molecules which the simulation may involve, thanks to improved algorithm performance.

Multicanonical hybrid Monte Carlo algorithm: Boosting simulations of compact QED

NASA Astrophysics Data System (ADS)

Arnold, G.; Schilling, K.; Lippert, Th.

1999-03-01

We demonstrate that substantial progress can be achieved in the study of the phase structure of four-dimensional compact QED by a joint use of hybrid Monte Carlo and multicanonical algorithms through an efficient parallel implementation. This is borne out by the observation of considerable speedup of tunnelling between the metastable states, close to the phase transition, on the Wilson line. We estimate that the creation of adequate samples (with order 100 flip-flops) becomes a matter of half a year's run time at 2 Gflops sustained performance for lattices of size up to 244.

A novel Monte Carlo algorithm for simulating crystals with McStas

NASA Astrophysics Data System (ADS)

Alianelli, L.; Sánchez del Río, M.; Felici, R.; Andersen, K. H.; Farhi, E.

2004-07-01

We developed an original Monte Carlo algorithm for the simulation of Bragg diffraction by mosaic, bent and gradient crystals. It has practical applications, as it can be used for simulating imperfect crystals (monochromators, analyzers and perhaps samples) in neutron ray-tracing packages, like McStas. The code we describe here provides a detailed description of the particle interaction with the microscopic homogeneous regions composing the crystal, therefore it can be used also for the calculation of quantities having a conceptual interest, as multiple scattering, or for the interpretation of experiments aiming at characterizing crystals, like diffraction topographs.

Probabilistic analysis algorithm for UA slope software program.

DOT National Transportation Integrated Search

2013-12-01

A reliability-based computational algorithm for using a single row and equally spaced drilled shafts to : stabilize an unstable slope has been developed in this research. The Monte-Carlo simulation (MCS) : technique was used in the previously develop...

Machine learning of frustrated classical spin models. I. Principal component analysis

NASA Astrophysics Data System (ADS)

Wang, Ce; Zhai, Hui

2017-10-01

This work aims at determining whether artificial intelligence can recognize a phase transition without prior human knowledge. If this were successful, it could be applied to, for instance, analyzing data from the quantum simulation of unsolved physical models. Toward this goal, we first need to apply the machine learning algorithm to well-understood models and see whether the outputs are consistent with our prior knowledge, which serves as the benchmark for this approach. In this work, we feed the computer data generated by the classical Monte Carlo simulation for the X Y model in frustrated triangular and union jack lattices, which has two order parameters and exhibits two phase transitions. We show that the outputs of the principal component analysis agree very well with our understanding of different orders in different phases, and the temperature dependences of the major components detect the nature and the locations of the phase transitions. Our work offers promise for using machine learning techniques to study sophisticated statistical models, and our results can be further improved by using principal component analysis with kernel tricks and the neural network method.

Estimating rare events in biochemical systems using conditional sampling.

PubMed

Sundar, V S

2017-01-28

The paper focuses on development of variance reduction strategies to estimate rare events in biochemical systems. Obtaining this probability using brute force Monte Carlo simulations in conjunction with the stochastic simulation algorithm (Gillespie's method) is computationally prohibitive. To circumvent this, important sampling tools such as the weighted stochastic simulation algorithm and the doubly weighted stochastic simulation algorithm have been proposed. However, these strategies require an additional step of determining the important region to sample from, which is not straightforward for most of the problems. In this paper, we apply the subset simulation method, developed as a variance reduction tool in the context of structural engineering, to the problem of rare event estimation in biochemical systems. The main idea is that the rare event probability is expressed as a product of more frequent conditional probabilities. These conditional probabilities are estimated with high accuracy using Monte Carlo simulations, specifically the Markov chain Monte Carlo method with the modified Metropolis-Hastings algorithm. Generating sample realizations of the state vector using the stochastic simulation algorithm is viewed as mapping the discrete-state continuous-time random process to the standard normal random variable vector. This viewpoint opens up the possibility of applying more sophisticated and efficient sampling schemes developed elsewhere to problems in stochastic chemical kinetics. The results obtained using the subset simulation method are compared with existing variance reduction strategies for a few benchmark problems, and a satisfactory improvement in computational time is demonstrated.

Parallel Monte Carlo Search for Hough Transform

NASA Astrophysics Data System (ADS)

Lopes, Raul H. C.; Franqueira, Virginia N. L.; Reid, Ivan D.; Hobson, Peter R.

2017-10-01

We investigate the problem of line detection in digital image processing and in special how state of the art algorithms behave in the presence of noise and whether CPU efficiency can be improved by the combination of a Monte Carlo Tree Search, hierarchical space decomposition, and parallel computing. The starting point of the investigation is the method introduced in 1962 by Paul Hough for detecting lines in binary images. Extended in the 1970s to the detection of space forms, what came to be known as Hough Transform (HT) has been proposed, for example, in the context of track fitting in the LHC ATLAS and CMS projects. The Hough Transform transfers the problem of line detection, for example, into one of optimization of the peak in a vote counting process for cells which contain the possible points of candidate lines. The detection algorithm can be computationally expensive both in the demands made upon the processor and on memory. Additionally, it can have a reduced effectiveness in detection in the presence of noise. Our first contribution consists in an evaluation of the use of a variation of the Radon Transform as a form of improving theeffectiveness of line detection in the presence of noise. Then, parallel algorithms for variations of the Hough Transform and the Radon Transform for line detection are introduced. An algorithm for Parallel Monte Carlo Search applied to line detection is also introduced. Their algorithmic complexities are discussed. Finally, implementations on multi-GPU and multicore architectures are discussed.

The Linked Neighbour List (LNL) method for fast off-lattice Monte Carlo simulations of fluids

NASA Astrophysics Data System (ADS)

Mazzeo, M. D.; Ricci, M.; Zannoni, C.

2010-03-01

We present a new algorithm, called linked neighbour list (LNL), useful to substantially speed up off-lattice Monte Carlo simulations of fluids by avoiding the computation of the molecular energy before every attempted move. We introduce a few variants of the LNL method targeted to minimise memory footprint or augment memory coherence and cache utilisation. Additionally, we present a few algorithms which drastically accelerate neighbour finding. We test our methods on the simulation of a dense off-lattice Gay-Berne fluid subjected to periodic boundary conditions observing a speedup factor of about 2.5 with respect to a well-coded implementation based on a conventional link-cell. We provide several implementation details of the different key data structures and algorithms used in this work.

Asteroid mass estimation with Markov-chain Monte Carlo

NASA Astrophysics Data System (ADS)

Siltala, L.; Granvik, M.

2017-09-01

We have developed a new Markov-chain Monte Carlo-based algorithm for asteroid mass estimation based on mutual encounters and tested it for several different asteroids. Our results are in line with previous literature values but suggest that uncertainties of prior estimates may be misleading as a consequence of using linearized methods.

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

Bolding, Simon R.; Cleveland, Mathew Allen; Morel, Jim E.

In this paper, we have implemented a new high-order low-order (HOLO) algorithm for solving thermal radiative transfer problems. The low-order (LO) system is based on the spatial and angular moments of the transport equation and a linear-discontinuous finite-element spatial representation, producing equations similar to the standard S 2 equations. The LO solver is fully implicit in time and efficiently resolves the nonlinear temperature dependence at each time step. The high-order (HO) solver utilizes exponentially convergent Monte Carlo (ECMC) to give a globally accurate solution for the angular intensity to a fixed-source pure-absorber transport problem. This global solution is used tomore » compute consistency terms, which require the HO and LO solutions to converge toward the same solution. The use of ECMC allows for the efficient reduction of statistical noise in the Monte Carlo solution, reducing inaccuracies introduced through the LO consistency terms. Finally, we compare results with an implicit Monte Carlo code for one-dimensional gray test problems and demonstrate the efficiency of ECMC over standard Monte Carlo in this HOLO algorithm.« less

An empirical approach to estimate near-infra-red photon propagation and optically induced drug release in brain tissues

NASA Astrophysics Data System (ADS)

Prabhu Verleker, Akshay; Fang, Qianqian; Choi, Mi-Ran; Clare, Susan; Stantz, Keith M.

2015-03-01

The purpose of this study is to develop an alternate empirical approach to estimate near-infra-red (NIR) photon propagation and quantify optically induced drug release in brain metastasis, without relying on computationally expensive Monte Carlo techniques (gold standard). Targeted drug delivery with optically induced drug release is a noninvasive means to treat cancers and metastasis. This study is part of a larger project to treat brain metastasis by delivering lapatinib-drug-nanocomplexes and activating NIR-induced drug release. The empirical model was developed using a weighted approach to estimate photon scattering in tissues and calibrated using a GPU based 3D Monte Carlo. The empirical model was developed and tested against Monte Carlo in optical brain phantoms for pencil beams (width 1mm) and broad beams (width 10mm). The empirical algorithm was tested against the Monte Carlo for different albedos along with diffusion equation and in simulated brain phantoms resembling white-matter (μs'=8.25mm-1, μa=0.005mm-1) and gray-matter (μs'=2.45mm-1, μa=0.035mm-1) at wavelength 800nm. The goodness of fit between the two models was determined using coefficient of determination (R-squared analysis). Preliminary results show the Empirical algorithm matches Monte Carlo simulated fluence over a wide range of albedo (0.7 to 0.99), while the diffusion equation fails for lower albedo. The photon fluence generated by empirical code matched the Monte Carlo in homogeneous phantoms (R2=0.99). While GPU based Monte Carlo achieved 300X acceleration compared to earlier CPU based models, the empirical code is 700X faster than the Monte Carlo for a typical super-Gaussian laser beam.

An Efficient MCMC Algorithm to Sample Binary Matrices with Fixed Marginals

ERIC Educational Resources Information Center

Verhelst, Norman D.

2008-01-01

Uniform sampling of binary matrices with fixed margins is known as a difficult problem. Two classes of algorithms to sample from a distribution not too different from the uniform are studied in the literature: importance sampling and Markov chain Monte Carlo (MCMC). Existing MCMC algorithms converge slowly, require a long burn-in period and yield…

SKIRT: The design of a suite of input models for Monte Carlo radiative transfer simulations

NASA Astrophysics Data System (ADS)

Baes, M.; Camps, P.

2015-09-01

The Monte Carlo method is the most popular technique to perform radiative transfer simulations in a general 3D geometry. The algorithms behind and acceleration techniques for Monte Carlo radiative transfer are discussed extensively in the literature, and many different Monte Carlo codes are publicly available. On the contrary, the design of a suite of components that can be used for the distribution of sources and sinks in radiative transfer codes has received very little attention. The availability of such models, with different degrees of complexity, has many benefits. For example, they can serve as toy models to test new physical ingredients, or as parameterised models for inverse radiative transfer fitting. For 3D Monte Carlo codes, this requires algorithms to efficiently generate random positions from 3D density distributions. We describe the design of a flexible suite of components for the Monte Carlo radiative transfer code SKIRT. The design is based on a combination of basic building blocks (which can be either analytical toy models or numerical models defined on grids or a set of particles) and the extensive use of decorators that combine and alter these building blocks to more complex structures. For a number of decorators, e.g. those that add spiral structure or clumpiness, we provide a detailed description of the algorithms that can be used to generate random positions. Advantages of this decorator-based design include code transparency, the avoidance of code duplication, and an increase in code maintainability. Moreover, since decorators can be chained without problems, very complex models can easily be constructed out of simple building blocks. Finally, based on a number of test simulations, we demonstrate that our design using customised random position generators is superior to a simpler design based on a generic black-box random position generator.

Wavelet Monte Carlo dynamics: A new algorithm for simulating the hydrodynamics of interacting Brownian particles

NASA Astrophysics Data System (ADS)

Dyer, Oliver T.; Ball, Robin C.

2017-03-01

We develop a new algorithm for the Brownian dynamics of soft matter systems that evolves time by spatially correlated Monte Carlo moves. The algorithm uses vector wavelets as its basic moves and produces hydrodynamics in the low Reynolds number regime propagated according to the Oseen tensor. When small moves are removed, the correlations closely approximate the Rotne-Prager tensor, itself widely used to correct for deficiencies in Oseen. We also include plane wave moves to provide the longest range correlations, which we detail for both infinite and periodic systems. The computational cost of the algorithm scales competitively with the number of particles simulated, N, scaling as N In N in homogeneous systems and as N in dilute systems. In comparisons to established lattice Boltzmann and Brownian dynamics algorithms, the wavelet method was found to be only a factor of order 1 times more expensive than the cheaper lattice Boltzmann algorithm in marginally semi-dilute simulations, while it is significantly faster than both algorithms at large N in dilute simulations. We also validate the algorithm by checking that it reproduces the correct dynamics and equilibrium properties of simple single polymer systems, as well as verifying the effect of periodicity on the mobility tensor.

Exact Dynamics via Poisson Process: a unifying Monte Carlo paradigm

NASA Astrophysics Data System (ADS)

Gubernatis, James

2014-03-01

A common computational task is solving a set of ordinary differential equations (o.d.e.'s). A little known theorem says that the solution of any set of o.d.e.'s is exactly solved by the expectation value over a set of arbitary Poisson processes of a particular function of the elements of the matrix that defines the o.d.e.'s. The theorem thus provides a new starting point to develop real and imaginary-time continous-time solvers for quantum Monte Carlo algorithms, and several simple observations enable various quantum Monte Carlo techniques and variance reduction methods to transfer to a new context. I will state the theorem, note a transformation to a very simple computational scheme, and illustrate the use of some techniques from the directed-loop algorithm in context of the wavefunction Monte Carlo method that is used to solve the Lindblad master equation for the dynamics of open quantum systems. I will end by noting that as the theorem does not depend on the source of the o.d.e.'s coming from quantum mechanics, it also enables the transfer of continuous-time methods from quantum Monte Carlo to the simulation of various classical equations of motion heretofore only solved deterministically.

Path-integral Monte Carlo method for Rényi entanglement entropies.

PubMed

Herdman, C M; Inglis, Stephen; Roy, P-N; Melko, R G; Del Maestro, A

2014-07-01

We introduce a quantum Monte Carlo algorithm to measure the Rényi entanglement entropies in systems of interacting bosons in the continuum. This approach is based on a path-integral ground state method that can be applied to interacting itinerant bosons in any spatial dimension with direct relevance to experimental systems of quantum fluids. We demonstrate how it may be used to compute spatial mode entanglement, particle partitioned entanglement, and the entanglement of particles, providing insights into quantum correlations generated by fluctuations, indistinguishability, and interactions. We present proof-of-principle calculations and benchmark against an exactly soluble model of interacting bosons in one spatial dimension. As this algorithm retains the fundamental polynomial scaling of quantum Monte Carlo when applied to sign-problem-free models, future applications should allow for the study of entanglement entropy in large-scale many-body systems of interacting bosons.

MUSiC - A general search for deviations from monte carlo predictions in CMS

NASA Astrophysics Data System (ADS)

Biallass, Philipp A.; CMS Collaboration

2009-06-01

A model independent analysis approach in CMS is presented, systematically scanning the data for deviations from the Monte Carlo expectation. Such an analysis can contribute to the understanding of the detector and the tuning of the event generators. Furthermore, due to the minimal theoretical bias this approach is sensitive to a variety of models of new physics, including those not yet thought of. Events are classified into event classes according to their particle content (muons, electrons, photons, jets and missing transverse energy). A broad scan of various distributions is performed, identifying significant deviations from the Monte Carlo simulation. The importance of systematic uncertainties is outlined, which are taken into account rigorously within the algorithm. Possible detector effects and generator issues, as well as models involving Supersymmetry and new heavy gauge bosons are used as an input to the search algorithm.

MUSiC - A Generic Search for Deviations from Monte Carlo Predictions in CMS

NASA Astrophysics Data System (ADS)

Hof, Carsten

2009-05-01

We present a model independent analysis approach, systematically scanning the data for deviations from the Standard Model Monte Carlo expectation. Such an analysis can contribute to the understanding of the CMS detector and the tuning of the event generators. Furthermore, due to the minimal theoretical bias this approach is sensitive to a variety of models of new physics, including those not yet thought of. Events are classified into event classes according to their particle content (muons, electrons, photons, jets and missing transverse energy). A broad scan of various distributions is performed, identifying significant deviations from the Monte Carlo simulation. We outline the importance of systematic uncertainties, which are taken into account rigorously within the algorithm. Possible detector effects and generator issues, as well as models involving supersymmetry and new heavy gauge bosons have been used as an input to the search algorithm.

A computational atomistic study of the relaxation of ion-bombarded c-Si on experimental time-scales: an application of the kinetic Activation Relaxation Technique

NASA Astrophysics Data System (ADS)

Béland, Laurent Karim; Mousseau, Normand

2012-02-01

The kinetic activation relaxation technique (kinetic ART) method, an off-lattice, self-learning kinetic Monte Carlo (KMC) algorithm with on-the-fly event search,ootnotetextL. K. B'eland, P. Brommer, F. El-Mellouhi, J.-F. Joly and N. Mousseau, Phys. Rev. E 84, 046704 (2011). is used to study the relaxation of c-Si after Si^- bombardment at 3 keV. We describe the evolution of the damaged areas at room-temperature and above for periods of the order of seconds, treating long-range elastic deformations exactly. We assess the stability of the nanoscale structures formed by the damage cascade and the mechanisms that govern post-implantation annealing.

Monte Carlo Transport for Electron Thermal Transport

NASA Astrophysics Data System (ADS)

Chenhall, Jeffrey; Cao, Duc; Moses, Gregory

2015-11-01

The iSNB (implicit Schurtz Nicolai Busquet multigroup electron thermal transport method of Cao et al. is adapted into a Monte Carlo transport method in order to better model the effects of non-local behavior. The end goal is a hybrid transport-diffusion method that combines Monte Carlo Transport with a discrete diffusion Monte Carlo (DDMC). The hybrid method will combine the efficiency of a diffusion method in short mean free path regions with the accuracy of a transport method in long mean free path regions. The Monte Carlo nature of the approach allows the algorithm to be massively parallelized. Work to date on the method will be presented. This work was supported by Sandia National Laboratory - Albuquerque and the University of Rochester Laboratory for Laser Energetics.

Learning reduced kinetic Monte Carlo models of complex chemistry from molecular dynamics.

PubMed

Yang, Qian; Sing-Long, Carlos A; Reed, Evan J

2017-08-01

We propose a novel statistical learning framework for automatically and efficiently building reduced kinetic Monte Carlo (KMC) models of large-scale elementary reaction networks from data generated by a single or few molecular dynamics simulations (MD). Existing approaches for identifying species and reactions from molecular dynamics typically use bond length and duration criteria, where bond duration is a fixed parameter motivated by an understanding of bond vibrational frequencies. In contrast, we show that for highly reactive systems, bond duration should be a model parameter that is chosen to maximize the predictive power of the resulting statistical model. We demonstrate our method on a high temperature, high pressure system of reacting liquid methane, and show that the learned KMC model is able to extrapolate more than an order of magnitude in time for key molecules. Additionally, our KMC model of elementary reactions enables us to isolate the most important set of reactions governing the behavior of key molecules found in the MD simulation. We develop a new data-driven algorithm to reduce the chemical reaction network which can be solved either as an integer program or efficiently using L1 regularization, and compare our results with simple count-based reduction. For our liquid methane system, we discover that rare reactions do not play a significant role in the system, and find that less than 7% of the approximately 2000 reactions observed from molecular dynamics are necessary to reproduce the molecular concentration over time of methane. The framework described in this work paves the way towards a genomic approach to studying complex chemical systems, where expensive MD simulation data can be reused to contribute to an increasingly large and accurate genome of elementary reactions and rates.

Learning reduced kinetic Monte Carlo models of complex chemistry from molecular dynamics

PubMed Central

Sing-Long, Carlos A.

2017-01-01

We propose a novel statistical learning framework for automatically and efficiently building reduced kinetic Monte Carlo (KMC) models of large-scale elementary reaction networks from data generated by a single or few molecular dynamics simulations (MD). Existing approaches for identifying species and reactions from molecular dynamics typically use bond length and duration criteria, where bond duration is a fixed parameter motivated by an understanding of bond vibrational frequencies. In contrast, we show that for highly reactive systems, bond duration should be a model parameter that is chosen to maximize the predictive power of the resulting statistical model. We demonstrate our method on a high temperature, high pressure system of reacting liquid methane, and show that the learned KMC model is able to extrapolate more than an order of magnitude in time for key molecules. Additionally, our KMC model of elementary reactions enables us to isolate the most important set of reactions governing the behavior of key molecules found in the MD simulation. We develop a new data-driven algorithm to reduce the chemical reaction network which can be solved either as an integer program or efficiently using L1 regularization, and compare our results with simple count-based reduction. For our liquid methane system, we discover that rare reactions do not play a significant role in the system, and find that less than 7% of the approximately 2000 reactions observed from molecular dynamics are necessary to reproduce the molecular concentration over time of methane. The framework described in this work paves the way towards a genomic approach to studying complex chemical systems, where expensive MD simulation data can be reused to contribute to an increasingly large and accurate genome of elementary reactions and rates. PMID:28989618

Learning reduced kinetic Monte Carlo models of complex chemistry from molecular dynamics

DOE PAGES

Yang, Qian; Sing-Long, Carlos A.; Reed, Evan J.

2017-06-19

Here, we propose a novel statistical learning framework for automatically and efficiently building reduced kinetic Monte Carlo (KMC) models of large-scale elementary reaction networks from data generated by a single or few molecular dynamics simulations (MD). Existing approaches for identifying species and reactions from molecular dynamics typically use bond length and duration criteria, where bond duration is a fixed parameter motivated by an understanding of bond vibrational frequencies. Conversely, we show that for highly reactive systems, bond duration should be a model parameter that is chosen to maximize the predictive power of the resulting statistical model. We demonstrate our methodmore » on a high temperature, high pressure system of reacting liquid methane, and show that the learned KMC model is able to extrapolate more than an order of magnitude in time for key molecules. Additionally, our KMC model of elementary reactions enables us to isolate the most important set of reactions governing the behavior of key molecules found in the MD simulation. We develop a new data-driven algorithm to reduce the chemical reaction network which can be solved either as an integer program or efficiently using L1 regularization, and compare our results with simple count-based reduction. For our liquid methane system, we discover that rare reactions do not play a significant role in the system, and find that less than 7% of the approximately 2000 reactions observed from molecular dynamics are necessary to reproduce the molecular concentration over time of methane. Furthermore, we describe a framework in this work that paves the way towards a genomic approach to studying complex chemical systems, where expensive MD simulation data can be reused to contribute to an increasingly large and accurate genome of elementary reactions and rates.« less

Instantons in Quantum Annealing: Thermally Assisted Tunneling Vs Quantum Monte Carlo Simulations

NASA Technical Reports Server (NTRS)

Jiang, Zhang; Smelyanskiy, Vadim N.; Boixo, Sergio; Isakov, Sergei V.; Neven, Hartmut; Mazzola, Guglielmo; Troyer, Matthias

2015-01-01

Recent numerical result (arXiv:1512.02206) from Google suggested that the D-Wave quantum annealer may have an asymptotic speed-up than simulated annealing, however, the asymptotic advantage disappears when it is compared to quantum Monte Carlo (a classical algorithm despite its name). We show analytically that the asymptotic scaling of quantum tunneling is exactly the same as the escape rate in quantum Monte Carlo for a class of problems. Thus, the Google result might be explained in our framework. We also found that the transition state in quantum Monte Carlo corresponds to the instanton solution in quantum tunneling problems, which is observed in numerical simulations.

Goal-oriented sensitivity analysis for lattice kinetic Monte Carlo simulations

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

Arampatzis, Georgios, E-mail: garab@math.uoc.gr; Department of Mathematics and Statistics, University of Massachusetts, Amherst, Massachusetts 01003; Katsoulakis, Markos A., E-mail: markos@math.umass.edu

2014-03-28

In this paper we propose a new class of coupling methods for the sensitivity analysis of high dimensional stochastic systems and in particular for lattice Kinetic Monte Carlo (KMC). Sensitivity analysis for stochastic systems is typically based on approximating continuous derivatives with respect to model parameters by the mean value of samples from a finite difference scheme. Instead of using independent samples the proposed algorithm reduces the variance of the estimator by developing a strongly correlated-“coupled”- stochastic process for both the perturbed and unperturbed stochastic processes, defined in a common state space. The novelty of our construction is that themore » new coupled process depends on the targeted observables, e.g., coverage, Hamiltonian, spatial correlations, surface roughness, etc., hence we refer to the proposed method as goal-oriented sensitivity analysis. In particular, the rates of the coupled Continuous Time Markov Chain are obtained as solutions to a goal-oriented optimization problem, depending on the observable of interest, by considering the minimization functional of the corresponding variance. We show that this functional can be used as a diagnostic tool for the design and evaluation of different classes of couplings. Furthermore, the resulting KMC sensitivity algorithm has an easy implementation that is based on the Bortz–Kalos–Lebowitz algorithm's philosophy, where events are divided in classes depending on level sets of the observable of interest. Finally, we demonstrate in several examples including adsorption, desorption, and diffusion Kinetic Monte Carlo that for the same confidence interval and observable, the proposed goal-oriented algorithm can be two orders of magnitude faster than existing coupling algorithms for spatial KMC such as the Common Random Number approach. We also provide a complete implementation of the proposed sensitivity analysis algorithms, including various spatial KMC examples, in a supplementary MATLAB source code.« less

Extraction of diffuse correlation spectroscopy flow index by integration of Nth-order linear model with Monte Carlo simulation

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

Shang, Yu; Lin, Yu; Yu, Guoqiang, E-mail: guoqiang.yu@uky.edu

2014-05-12

Conventional semi-infinite solution for extracting blood flow index (BFI) from diffuse correlation spectroscopy (DCS) measurements may cause errors in estimation of BFI (αD{sub B}) in tissues with small volume and large curvature. We proposed an algorithm integrating Nth-order linear model of autocorrelation function with the Monte Carlo simulation of photon migrations in tissue for the extraction of αD{sub B}. The volume and geometry of the measured tissue were incorporated in the Monte Carlo simulation, which overcome the semi-infinite restrictions. The algorithm was tested using computer simulations on four tissue models with varied volumes/geometries and applied on an in vivo strokemore » model of mouse. Computer simulations shows that the high-order (N ≥ 5) linear algorithm was more accurate in extracting αD{sub B} (errors < ±2%) from the noise-free DCS data than the semi-infinite solution (errors: −5.3% to −18.0%) for different tissue models. Although adding random noises to DCS data resulted in αD{sub B} variations, the mean values of errors in extracting αD{sub B} were similar to those reconstructed from the noise-free DCS data. In addition, the errors in extracting the relative changes of αD{sub B} using both linear algorithm and semi-infinite solution were fairly small (errors < ±2.0%) and did not rely on the tissue volume/geometry. The experimental results from the in vivo stroke mice agreed with those in simulations, demonstrating the robustness of the linear algorithm. DCS with the high-order linear algorithm shows the potential for the inter-subject comparison and longitudinal monitoring of absolute BFI in a variety of tissues/organs with different volumes/geometries.« less

An Improved Elastic and Nonelastic Neutron Transport Algorithm for Space Radiation

NASA Technical Reports Server (NTRS)

Clowdsley, Martha S.; Wilson, John W.; Heinbockel, John H.; Tripathi, R. K.; Singleterry, Robert C., Jr.; Shinn, Judy L.

2000-01-01

A neutron transport algorithm including both elastic and nonelastic particle interaction processes for use in space radiation protection for arbitrary shield material is developed. The algorithm is based upon a multiple energy grouping and analysis of the straight-ahead Boltzmann equation by using a mean value theorem for integrals. The algorithm is then coupled to the Langley HZETRN code through a bidirectional neutron evaporation source term. Evaluation of the neutron fluence generated by the solar particle event of February 23, 1956, for an aluminum water shield-target configuration is then compared with MCNPX and LAHET Monte Carlo calculations for the same shield-target configuration. With the Monte Carlo calculation as a benchmark, the algorithm developed in this paper showed a great improvement in results over the unmodified HZETRN solution. In addition, a high-energy bidirectional neutron source based on a formula by Ranft showed even further improvement of the fluence results over previous results near the front of the water target where diffusion out the front surface is important. Effects of improved interaction cross sections are modest compared with the addition of the high-energy bidirectional source terms.

Delving Into Dissipative Quantum Dynamics: From Approximate to Numerically Exact Approaches

NASA Astrophysics Data System (ADS)

Chen, Hsing-Ta

In this thesis, I explore dissipative quantum dynamics of several prototypical model systems via various approaches, ranging from approximate to numerically exact schemes. In particular, in the realm of the approximate I explore the accuracy of Pade-resummed master equations and the fewest switches surface hopping (FSSH) algorithm for the spin-boson model, and non-crossing approximations (NCA) for the Anderson-Holstein model. Next, I develop new and exact Monte Carlo approaches and test them on the spin-boson model. I propose well-defined criteria for assessing the accuracy of Pade-resummed quantum master equations, which correctly demarcate the regions of parameter space where the Pade approximation is reliable. I continue the investigation of spin-boson dynamics by benchmark comparisons of the semiclassical FSSH algorithm to exact dynamics over a wide range of parameters. Despite small deviations from golden-rule scaling in the Marcus regime, standard surface hopping algorithm is found to be accurate over a large portion of parameter space. The inclusion of decoherence corrections via the augmented FSSH algorithm improves the accuracy of dynamical behavior compared to exact simulations, but the effects are generally not dramatic for the cases I consider. Next, I introduce new methods for numerically exact real-time simulation based on real-time diagrammatic Quantum Monte Carlo (dQMC) and the inchworm algorithm. These methods optimally recycle Monte Carlo information from earlier times to greatly suppress the dynamical sign problem. In the context of the spin-boson model, I formulate the inchworm expansion in two distinct ways: the first with respect to an expansion in the system-bath coupling and the second as an expansion in the diabatic coupling. In addition, a cumulant version of the inchworm Monte Carlo method is motivated by the latter expansion, which allows for further suppression of the growth of the sign error. I provide a comprehensive comparison of the performance of the inchworm Monte Carlo algorithms to other exact methodologies as well as a discussion of the relative advantages and disadvantages of each. Finally, I investigate the dynamical interplay between the electron-electron interaction and the electron-phonon coupling within the Anderson-Holstein model via two complementary NCAs: the first is constructed around the weak-coupling limit and the second around the polaron limit. The influence of phonons on spectral and transport properties is explored in equilibrium, for non-equilibrium steady state and for transient dynamics after a quench. I find the two NCAs disagree in nontrivial ways, indicating that more reliable approaches to the problem are needed. The complementary frameworks used here pave the way for numerically exact methods based on inchworm dQMC algorithms capable of treating open systems simultaneously coupled to multiple fermionic and bosonic baths.

Monte Carlo calculation of large and small-angle electron scattering in air

NASA Astrophysics Data System (ADS)

Cohen, B. I.; Higginson, D. P.; Eng, C. D.; Farmer, W. A.; Friedman, A.; Grote, D. P.; Larson, D. J.

2017-11-01

A Monte Carlo method for angle scattering of electrons in air that accommodates the small-angle multiple scattering and larger-angle single scattering limits is introduced. The algorithm is designed for use in a particle-in-cell simulation of electron transport and electromagnetic wave effects in air. The method is illustrated in example calculations.

Statistical hadronization and microcanonical ensemble

DOE PAGES

Becattini, F.; Ferroni, L.

2004-01-01

We present a Monte Carlo calculation of the microcanonical ensemble of the of the ideal hadron-resonance gas including all known states up to a mass of 1. 8 GeV, taking into account quantum statistics. The computing method is a development of a previous one based on a Metropolis Monte Carlo algorithm, with a the grand-canonical limit of the multi-species multiplicity distribution as proposal matrix. The microcanonical average multiplicities of the various hadron species are found to converge to the canonical ones for moderately low values of the total energy. This algorithm opens the way for event generators based for themore » statistical hadronization model.« less

Experimental evaluation of a GPU-based Monte Carlo dose calculation algorithm in the Monaco treatment planning system.

PubMed

Paudel, Moti R; Kim, Anthony; Sarfehnia, Arman; Ahmad, Sayed B; Beachey, David J; Sahgal, Arjun; Keller, Brian M

2016-11-08

A new GPU-based Monte Carlo dose calculation algorithm (GPUMCD), devel-oped by the vendor Elekta for the Monaco treatment planning system (TPS), is capable of modeling dose for both a standard linear accelerator and an Elekta MRI linear accelerator. We have experimentally evaluated this algorithm for a standard Elekta Agility linear accelerator. A beam model was developed in the Monaco TPS (research version 5.09.06) using the commissioned beam data for a 6 MV Agility linac. A heterogeneous phantom representing several scenarios - tumor-in-lung, lung, and bone-in-tissue - was designed and built. Dose calculations in Monaco were done using both the current clinical Monte Carlo algorithm, XVMC, and the new GPUMCD algorithm. Dose calculations in a Pinnacle TPS were also produced using the collapsed cone convolution (CCC) algorithm with heterogeneity correc-tion. Calculations were compared with the measured doses using an ionization chamber (A1SL) and Gafchromic EBT3 films for 2 × 2 cm2, 5 × 5 cm2, and 10 × 10 cm2 field sizes. The percentage depth doses (PDDs) calculated by XVMC and GPUMCD in a homogeneous solid water phantom were within 2%/2 mm of film measurements and within 1% of ion chamber measurements. For the tumor-in-lung phantom, the calculated doses were within 2.5%/2.5 mm of film measurements for GPUMCD. For the lung phantom, doses calculated by all of the algorithms were within 3%/3 mm of film measurements, except for the 2 × 2 cm2 field size where the CCC algorithm underestimated the depth dose by ~ 5% in a larger extent of the lung region. For the bone phantom, all of the algorithms were equivalent and calculated dose to within 2%/2 mm of film measurements, except at the interfaces. Both GPUMCD and XVMC showed interface effects, which were more pronounced for GPUMCD and were comparable to film measurements, whereas the CCC algorithm showed these effects poorly. © 2016 The Authors.

A clinical study of lung cancer dose calculation accuracy with Monte Carlo simulation.

PubMed

Zhao, Yanqun; Qi, Guohai; Yin, Gang; Wang, Xianliang; Wang, Pei; Li, Jian; Xiao, Mingyong; Li, Jie; Kang, Shengwei; Liao, Xiongfei

2014-12-16

The accuracy of dose calculation is crucial to the quality of treatment planning and, consequently, to the dose delivered to patients undergoing radiation therapy. Current general calculation algorithms such as Pencil Beam Convolution (PBC) and Collapsed Cone Convolution (CCC) have shortcomings in regard to severe inhomogeneities, particularly in those regions where charged particle equilibrium does not hold. The aim of this study was to evaluate the accuracy of the PBC and CCC algorithms in lung cancer radiotherapy using Monte Carlo (MC) technology. Four treatment plans were designed using Oncentra Masterplan TPS for each patient. Two intensity-modulated radiation therapy (IMRT) plans were developed using the PBC and CCC algorithms, and two three-dimensional conformal therapy (3DCRT) plans were developed using the PBC and CCC algorithms. The DICOM-RT files of the treatment plans were exported to the Monte Carlo system to recalculate. The dose distributions of GTV, PTV and ipsilateral lung calculated by the TPS and MC were compared. For 3DCRT and IMRT plans, the mean dose differences for GTV between the CCC and MC increased with decreasing of the GTV volume. For IMRT, the mean dose differences were found to be higher than that of 3DCRT. The CCC algorithm overestimated the GTV mean dose by approximately 3% for IMRT. For 3DCRT plans, when the volume of the GTV was greater than 100 cm(3), the mean doses calculated by CCC and MC almost have no difference. PBC shows large deviations from the MC algorithm. For the dose to the ipsilateral lung, the CCC algorithm overestimated the dose to the entire lung, and the PBC algorithm overestimated V20 but underestimated V5; the difference in V10 was not statistically significant. PBC substantially overestimates the dose to the tumour, but the CCC is similar to the MC simulation. It is recommended that the treatment plans for lung cancer be developed using an advanced dose calculation algorithm other than PBC. MC can accurately calculate the dose distribution in lung cancer and can provide a notably effective tool for benchmarking the performance of other dose calculation algorithms within patients.

Water Oxidation Catalysis for NiOOH by a Metropolis Monte Carlo Algorithm.

PubMed

Hareli, Chen; Caspary Toroker, Maytal

2018-05-08

Understanding catalytic mechanisms is important for discovering better catalysts, particularly for water splitting reactions that are of great interest to the renewable energy field. One of the best performing catalysts for water oxidation is nickel oxyhydroxide (NiOOH). However, only one mechanism has been adopted so far for modeling catalysis of the active plane: β-NiOOH(01̅5). In order to understand how a second reaction mechanism affects catalysis, we perform Density Functional Theory + U (DFT+U) calculations of a second mechanism for water oxidation reaction of NiOOH. Then, we use a Metropolis Monte Carlo algorithm to calculate how many catalytic cycles are completed when two reaction mechanisms are competing. We find that within the Metropolis algorithm, the second mechanism has a higher overpotential and is therefore not active even for large applied biases.

Improved cache performance in Monte Carlo transport calculations using energy banding

NASA Astrophysics Data System (ADS)

Siegel, A.; Smith, K.; Felker, K.; Romano, P.; Forget, B.; Beckman, P.

2014-04-01

We present an energy banding algorithm for Monte Carlo (MC) neutral particle transport simulations which depend on large cross section lookup tables. In MC codes, read-only cross section data tables are accessed frequently, exhibit poor locality, and are typically too much large to fit in fast memory. Thus, performance is often limited by long latencies to RAM, or by off-node communication latencies when the data footprint is very large and must be decomposed on a distributed memory machine. The proposed energy banding algorithm allows maximal temporal reuse of data in band sizes that can flexibly accommodate different architectural features. The energy banding algorithm is general and has a number of benefits compared to the traditional approach. In the present analysis we explore its potential to achieve improvements in time-to-solution on modern cache-based architectures.

Object-Oriented/Data-Oriented Design of a Direct Simulation Monte Carlo Algorithm

NASA Technical Reports Server (NTRS)

Liechty, Derek S.

2014-01-01

Over the past decade, there has been much progress towards improved phenomenological modeling and algorithmic updates for the direct simulation Monte Carlo (DSMC) method, which provides a probabilistic physical simulation of gas Rows. These improvements have largely been based on the work of the originator of the DSMC method, Graeme Bird. Of primary importance are improved chemistry, internal energy, and physics modeling and a reduction in time to solution. These allow for an expanded range of possible solutions In altitude and velocity space. NASA's current production code, the DSMC Analysis Code (DAC), is well-established and based on Bird's 1994 algorithms written in Fortran 77 and has proven difficult to upgrade. A new DSMC code is being developed in the C++ programming language using object-oriented and data-oriented design paradigms to facilitate the inclusion of the recent improvements and future development activities. The development efforts on the new code, the Multiphysics Algorithm with Particles (MAP), are described, and performance comparisons are made with DAC.

A global reaction route mapping-based kinetic Monte Carlo algorithm

NASA Astrophysics Data System (ADS)

Mitchell, Izaac; Irle, Stephan; Page, Alister J.

2016-07-01

We propose a new on-the-fly kinetic Monte Carlo (KMC) method that is based on exhaustive potential energy surface searching carried out with the global reaction route mapping (GRRM) algorithm. Starting from any given equilibrium state, this GRRM-KMC algorithm performs a one-step GRRM search to identify all surrounding transition states. Intrinsic reaction coordinate pathways are then calculated to identify potential subsequent equilibrium states. Harmonic transition state theory is used to calculate rate constants for all potential pathways, before a standard KMC accept/reject selection is performed. The selected pathway is then used to propagate the system forward in time, which is calculated on the basis of 1st order kinetics. The GRRM-KMC algorithm is validated here in two challenging contexts: intramolecular proton transfer in malonaldehyde and surface carbon diffusion on an iron nanoparticle. We demonstrate that in both cases the GRRM-KMC method is capable of reproducing the 1st order kinetics observed during independent quantum chemical molecular dynamics simulations using the density-functional tight-binding potential.

Efficiency in nonequilibrium molecular dynamics Monte Carlo simulations

DOE PAGES

Radak, Brian K.; Roux, Benoît

2016-10-07

Hybrid algorithms combining nonequilibrium molecular dynamics and Monte Carlo (neMD/MC) offer a powerful avenue for improving the sampling efficiency of computer simulations of complex systems. These neMD/MC algorithms are also increasingly finding use in applications where conventional approaches are impractical, such as constant-pH simulations with explicit solvent. However, selecting an optimal nonequilibrium protocol for maximum efficiency often represents a non-trivial challenge. This work evaluates the efficiency of a broad class of neMD/MC algorithms and protocols within the theoretical framework of linear response theory. The approximations are validated against constant pH-MD simulations and shown to provide accurate predictions of neMD/MC performance.more » An assessment of a large set of protocols confirms (both theoretically and empirically) that a linear work protocol gives the best neMD/MC performance. Lastly, a well-defined criterion for optimizing the time parameters of the protocol is proposed and demonstrated with an adaptive algorithm that improves the performance on-the-fly with minimal cost.« less

A global reaction route mapping-based kinetic Monte Carlo algorithm.

PubMed

Mitchell, Izaac; Irle, Stephan; Page, Alister J

2016-07-14

We propose a new on-the-fly kinetic Monte Carlo (KMC) method that is based on exhaustive potential energy surface searching carried out with the global reaction route mapping (GRRM) algorithm. Starting from any given equilibrium state, this GRRM-KMC algorithm performs a one-step GRRM search to identify all surrounding transition states. Intrinsic reaction coordinate pathways are then calculated to identify potential subsequent equilibrium states. Harmonic transition state theory is used to calculate rate constants for all potential pathways, before a standard KMC accept/reject selection is performed. The selected pathway is then used to propagate the system forward in time, which is calculated on the basis of 1st order kinetics. The GRRM-KMC algorithm is validated here in two challenging contexts: intramolecular proton transfer in malonaldehyde and surface carbon diffusion on an iron nanoparticle. We demonstrate that in both cases the GRRM-KMC method is capable of reproducing the 1st order kinetics observed during independent quantum chemical molecular dynamics simulations using the density-functional tight-binding potential.

Application of genetic algorithms to focal mechanism determination

NASA Astrophysics Data System (ADS)

Kobayashi, Reiji; Nakanishi, Ichiro

1994-04-01

Genetic algorithms are a new class of methods for global optimization. They resemble Monte Carlo techniques, but search for solutions more efficiently than uniform Monte Carlo sampling. In the field of geophysics, genetic algorithms have recently been used to solve some non-linear inverse problems (e.g., earthquake location, waveform inversion, migration velocity estimation). We present an application of genetic algorithms to focal mechanism determination from first-motion polarities of P-waves and apply our method to two recent large events, the Kushiro-oki earthquake of January 15, 1993 and the SW Hokkaido (Japan Sea) earthquake of July 12, 1993. Initial solution and curvature information of the objective function that gradient methods need are not required in our approach. Moreover globally optimal solutions can be efficiently obtained. Calculation of polarities based on double-couple models is the most time-consuming part of the source mechanism determination. The amount of calculations required by the method designed in this study is much less than that of previous grid search methods.

Teaching Monte Carlo Strategies for Earth System Modelling using a Guided Group-Learning Approach in the Classroom

NASA Astrophysics Data System (ADS)

Wagener, T.; Pianosi, F.; Woods, R. A.

2016-12-01

The need for quantifying uncertainty in earth system modelling has now been well established on both scientific and policy-making grounds. There is an urgent need to bring the skills and tools needed for doing so into practice. However, such topics are currently largely constrained to specialist graduate courses or to short courses for PhD students. Teaching the advanced skills needed for implementing and for using uncertainty analysis is difficult because students feel that it is inaccessible and it can be boring if presented using frontal teaching in the classroom. While we have made significant advancement in sharing teaching material, sometimes even including teaching notes (Wagener et al., 2012, Hydrology and Earth System Sciences), there is great need for understanding how we can bring such advanced topics into the undergraduate (and even graduate) curriculum in an effective manner. We present the results of our efforts to teach Matlab-based tools for uncertainty quantification in earth system modelling in a civil engineering undergraduate course. We use the example of teaching Monte Carlo strategies, the basis for the most widely used uncertainty quantification approaches, through the use of guided group-learning activities in the classroom. We utilize a three-step approach: [1] basic introduction to the problem, [2] guided group-learning to develop a possible solution, [3] comparison of possible solutions with state-of-the-art algorithms across groups. Our initial testing in an undergraduate course suggests that (i) overall students find a group-learning approach more engaging, (ii) that different students take charge of advancing the discussion at different stages or for different problems, and (iii) that making appropriate suggestions (facilitator) to guide the discussion keeps the speed of advancement sufficiently high. We present the approach, our initial results and suggest how a wider course on earth system modelling could be formulated in this manner.

Dose calculations using artificial neural networks: A feasibility study for photon beams

NASA Astrophysics Data System (ADS)

Vasseur, Aurélien; Makovicka, Libor; Martin, Éric; Sauget, Marc; Contassot-Vivier, Sylvain; Bahi, Jacques

2008-04-01

Direct dose calculations are a crucial requirement for Treatment Planning Systems. Some methods, such as Monte Carlo, explicitly model particle transport, others depend upon tabulated data or analytic formulae. However, their computation time is too lengthy for clinical use, or accuracy is insufficient, especially for recent techniques such as Intensity-Modulated Radiotherapy. Based on artificial neural networks (ANNs), a new solution is proposed and this work extends the properties of such an algorithm and is called NeuRad. Prior to any calculations, a first phase known as the learning process is necessary. Monte Carlo dose distributions in homogeneous media are used, and the ANN is then acquired. According to the training base, it can be used as a dose engine for either heterogeneous media or for an unknown material. In this report, two networks were created in order to compute dose distribution within a homogeneous phantom made of an unknown material and within an inhomogeneous phantom made of water and TA6V4 (titanium alloy corresponding to hip prosthesis). All NeuRad results were compared to Monte Carlo distributions. The latter required about 7 h on a dedicated cluster (10 nodes). NeuRad learning requires between 8 and 18 h (depending upon the size of the training base) on a single low-end computer. However, the results of dose computation with the ANN are available in less than 2 s, again using a low-end computer, for a 150×1×150 voxels phantom. In the case of homogeneous medium, the mean deviation in the high dose region was less than 1.7%. With a TA6V4 hip prosthesis bathed in water, the mean deviation in the high dose region was less than 4.1%. Further improvements in NeuRad will have to include full 3D calculations, inhomogeneity management and input definitions.

SU-E-T-202: Impact of Monte Carlo Dose Calculation Algorithm On Prostate SBRT Treatments

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

Venencia, C; Garrigo, E; Cardenas, J

2014-06-01

Purpose: The purpose of this work was to quantify the dosimetric impact of using Monte Carlo algorithm on pre calculated SBRT prostate treatment with pencil beam dose calculation algorithm. Methods: A 6MV photon beam produced by a Novalis TX (BrainLAB-Varian) linear accelerator equipped with HDMLC was used. Treatment plans were done using 9 fields with Iplanv4.5 (BrainLAB) and dynamic IMRT modality. Institutional SBRT protocol uses a total dose to the prostate of 40Gy in 5 fractions, every other day. Dose calculation is done by pencil beam (2mm dose resolution), heterogeneity correction and dose volume constraint (UCLA) for PTV D95%=40Gy andmore » D98%>39.2Gy, Rectum V20Gy<50%, V32Gy<20%, V36Gy<10% and V40Gy<5%, Bladder V20Gy<40% and V40Gy<10%, femoral heads V16Gy<5%, penile bulb V25Gy<3cc, urethra and overlap region between PTV and PRV Rectum Dmax<42Gy. 10 SBRT treatments plans were selected and recalculated using Monte Carlo with 2mm spatial resolution and mean variance of 2%. DVH comparisons between plans were done. Results: The average difference between PTV doses constraints were within 2%. However 3 plans have differences higher than 3% which does not meet the D98% criteria (>39.2Gy) and should have been renormalized. Dose volume constraint differences for rectum, bladder, femoral heads and penile bulb were les than 2% and within tolerances. Urethra region and overlapping between PTV and PRV Rectum shows increment of dose in all plans. The average difference for urethra region was 2.1% with a maximum of 7.8% and for the overlapping region 2.5% with a maximum of 8.7%. Conclusion: Monte Carlo dose calculation on dynamic IMRT treatments could affects on plan normalization. Dose increment in critical region of urethra and PTV overlapping region with PTV could have clinical consequences which need to be studied. The use of Monte Carlo dose calculation algorithm is limited because inverse planning dose optimization use only pencil beam.« less

Monte Carlo calculation of large and small-angle electron scattering in air

DOE PAGES

Cohen, B. I.; Higginson, D. P.; Eng, C. D.; ...

2017-08-12

A Monte Carlo method for angle scattering of electrons in air that accommodates the small-angle multiple scattering and larger-angle single scattering limits is introduced. In this work, the algorithm is designed for use in a particle-in-cell simulation of electron transport and electromagnetic wave effects in air. The method is illustrated in example calculations.

Monte Carlo method for magnetic impurities in metals

NASA Technical Reports Server (NTRS)

Hirsch, J. E.; Fye, R. M.

1986-01-01

The paper discusses a Monte Carlo algorithm to study properties of dilute magnetic alloys; the method can treat a small number of magnetic impurities interacting wiith the conduction electrons in a metal. Results for the susceptibility of a single Anderson impurity in the symmetric case show the expected universal behavior at low temperatures. Some results for two Anderson impurities are also discussed.

An NCME Instructional Module on Estimating Item Response Theory Models Using Markov Chain Monte Carlo Methods

ERIC Educational Resources Information Center

Kim, Jee-Seon; Bolt, Daniel M.

2007-01-01

The purpose of this ITEMS module is to provide an introduction to Markov chain Monte Carlo (MCMC) estimation for item response models. A brief description of Bayesian inference is followed by an overview of the various facets of MCMC algorithms, including discussion of prior specification, sampling procedures, and methods for evaluating chain…

Modeling 2D and 3D diffusion.

PubMed

Saxton, Michael J

2007-01-01

Modeling obstructed diffusion is essential to the understanding of diffusion-mediated processes in the crowded cellular environment. Simple Monte Carlo techniques for modeling obstructed random walks are explained and related to Brownian dynamics and more complicated Monte Carlo methods. Random number generation is reviewed in the context of random walk simulations. Programming techniques and event-driven algorithms are discussed as ways to speed simulations.

Simulation and performance of an artificial retina for 40 MHz track reconstruction

DOE PAGES

Abba, A.; Bedeschi, F.; Citterio, M.; ...

2015-03-05

We present the results of a detailed simulation of the artificial retina pattern-recognition algorithm, designed to reconstruct events with hundreds of charged-particle tracks in pixel and silicon detectors at LHCb with LHC crossing frequency of 40 MHz. Performances of the artificial retina algorithm are assessed using the official Monte Carlo samples of the LHCb experiment. We found performances for the retina pattern-recognition algorithm comparable with the full LHCb reconstruction algorithm.

Testing algorithms for critical slowing down

NASA Astrophysics Data System (ADS)

Cossu, Guido; Boyle, Peter; Christ, Norman; Jung, Chulwoo; Jüttner, Andreas; Sanfilippo, Francesco

2018-03-01

We present the preliminary tests on two modifications of the Hybrid Monte Carlo (HMC) algorithm. Both algorithms are designed to travel much farther in the Hamiltonian phase space for each trajectory and reduce the autocorrelations among physical observables thus tackling the critical slowing down towards the continuum limit. We present a comparison of costs of the new algorithms with the standard HMC evolution for pure gauge fields, studying the autocorrelation times for various quantities including the topological charge.

How Monte Carlo heuristics aid to identify the physical processes of drug release kinetics.

PubMed

Lecca, Paola

2018-01-01

We implement a Monte Carlo heuristic algorithm to model drug release from a solid dosage form. We show that with Monte Carlo simulations it is possible to identify and explain the causes of the unsatisfactory predictive power of current drug release models. It is well known that the power-law, the exponential models, as well as those derived from or inspired by them accurately reproduce only the first 60% of the release curve of a drug from a dosage form. In this study, by using Monte Carlo simulation approaches, we show that these models fit quite accurately almost the entire release profile when the release kinetics is not governed by the coexistence of different physico-chemical mechanisms. We show that the accuracy of the traditional models are comparable with those of Monte Carlo heuristics when these heuristics approximate and oversimply the phenomenology of drug release. This observation suggests to develop and use novel Monte Carlo simulation heuristics able to describe the complexity of the release kinetics, and consequently to generate data more similar to those observed in real experiments. Implementing Monte Carlo simulation heuristics of the drug release phenomenology may be much straightforward and efficient than hypothesizing and implementing from scratch complex mathematical models of the physical processes involved in drug release. Identifying and understanding through simulation heuristics what processes of this phenomenology reproduce the observed data and then formalize them in mathematics may allow avoiding time-consuming, trial-error based regression procedures. Three bullet points, highlighting the customization of the procedure. •An efficient heuristics based on Monte Carlo methods for simulating drug release from solid dosage form encodes is presented. It specifies the model of the physical process in a simple but accurate way in the formula of the Monte Carlo Micro Step (MCS) time interval.•Given the experimentally observed curve of drug release, we point out how Monte Carlo heuristics can be integrated in an evolutionary algorithmic approach to infer the mode of MCS best fitting the observed data, and thus the observed release kinetics.•The software implementing the method is written in R language, the free most used language in the bioinformaticians community.

Improved algorithm for estimating optical properties of food and biological materials using spatially-resolved diffuse reflectance

USDA-ARS?s Scientific Manuscript database

In this research, the inverse algorithm for estimating optical properties of food and biological materials from spatially-resolved diffuse reflectance was optimized in terms of data smoothing, normalization and spatial region of reflectance profile for curve fitting. Monte Carlo simulation was used ...

RNA inverse folding using Monte Carlo tree search.

PubMed

Yang, Xiufeng; Yoshizoe, Kazuki; Taneda, Akito; Tsuda, Koji

2017-11-06

Artificially synthesized RNA molecules provide important ways for creating a variety of novel functional molecules. State-of-the-art RNA inverse folding algorithms can design simple and short RNA sequences of specific GC content, that fold into the target RNA structure. However, their performance is not satisfactory in complicated cases. We present a new inverse folding algorithm called MCTS-RNA, which uses Monte Carlo tree search (MCTS), a technique that has shown exceptional performance in Computer Go recently, to represent and discover the essential part of the sequence space. To obtain high accuracy, initial sequences generated by MCTS are further improved by a series of local updates. Our algorithm has an ability to control the GC content precisely and can deal with pseudoknot structures. Using common benchmark datasets for evaluation, MCTS-RNA showed a lot of promise as a standard method of RNA inverse folding. MCTS-RNA is available at https://github.com/tsudalab/MCTS-RNA .

Modified Monte Carlo method for study of electron transport in degenerate electron gas in the presence of electron-electron interactions, application to graphene

NASA Astrophysics Data System (ADS)

Borowik, Piotr; Thobel, Jean-Luc; Adamowicz, Leszek

2017-07-01

Standard computational methods used to take account of the Pauli Exclusion Principle into Monte Carlo (MC) simulations of electron transport in semiconductors may give unphysical results in low field regime, where obtained electron distribution function takes values exceeding unity. Modified algorithms were already proposed and allow to correctly account for electron scattering on phonons or impurities. Present paper extends this approach and proposes improved simulation scheme allowing including Pauli exclusion principle for electron-electron (e-e) scattering into MC simulations. Simulations with significantly reduced computational cost recreate correct values of the electron distribution function. Proposed algorithm is applied to study transport properties of degenerate electrons in graphene with e-e interactions. This required adapting the treatment of e-e scattering in the case of linear band dispersion relation. Hence, this part of the simulation algorithm is described in details.

ALGEBRA: ALgorithm for the heterogeneous dosimetry based on GEANT4 for BRAchytherapy.

PubMed

Afsharpour, H; Landry, G; D'Amours, M; Enger, S; Reniers, B; Poon, E; Carrier, J-F; Verhaegen, F; Beaulieu, L

2012-06-07

Task group 43 (TG43)-based dosimetry algorithms are efficient for brachytherapy dose calculation in water. However, human tissues have chemical compositions and densities different than water. Moreover, the mutual shielding effect of seeds on each other (interseed attenuation) is neglected in the TG43-based dosimetry platforms. The scientific community has expressed the need for an accurate dosimetry platform in brachytherapy. The purpose of this paper is to present ALGEBRA, a Monte Carlo platform for dosimetry in brachytherapy which is sufficiently fast and accurate for clinical and research purposes. ALGEBRA is based on the GEANT4 Monte Carlo code and is capable of handling the DICOM RT standard to recreate a virtual model of the treated site. Here, the performance of ALGEBRA is presented for the special case of LDR brachytherapy in permanent prostate and breast seed implants. However, the algorithm is also capable of handling other treatments such as HDR brachytherapy.

Communication: Calculation of interatomic forces and optimization of molecular geometry with auxiliary-field quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Motta, Mario; Zhang, Shiwei

2018-05-01

We propose an algorithm for accurate, systematic, and scalable computation of interatomic forces within the auxiliary-field quantum Monte Carlo (AFQMC) method. The algorithm relies on the Hellmann-Feynman theorem and incorporates Pulay corrections in the presence of atomic orbital basis sets. We benchmark the method for small molecules by comparing the computed forces with the derivatives of the AFQMC potential energy surface and by direct comparison with other quantum chemistry methods. We then perform geometry optimizations using the steepest descent algorithm in larger molecules. With realistic basis sets, we obtain equilibrium geometries in agreement, within statistical error bars, with experimental values. The increase in computational cost for computing forces in this approach is only a small prefactor over that of calculating the total energy. This paves the way for a general and efficient approach for geometry optimization and molecular dynamics within AFQMC.

A Comparison of Experimental EPMA Data and Monte Carlo Simulations

NASA Technical Reports Server (NTRS)

Carpenter, P. K.

2004-01-01

Monte Carlo (MC) modeling shows excellent prospects for simulating electron scattering and x-ray emission from complex geometries, and can be compared to experimental measurements using electron-probe microanalysis (EPMA) and phi(rho z) correction algorithms. Experimental EPMA measurements made on NIST SRM 481 (AgAu) and 482 (CuAu) alloys, at a range of accelerating potential and instrument take-off angles, represent a formal microanalysis data set that has been used to develop phi(rho z) correction algorithms. The accuracy of MC calculations obtained using the NIST, WinCasino, WinXray, and Penelope MC packages will be evaluated relative to these experimental data. There is additional information contained in the extended abstract.

Rényi information flow in the Ising model with single-spin dynamics.

PubMed

Deng, Zehui; Wu, Jinshan; Guo, Wenan

2014-12-01

The n-index Rényi mutual information and transfer entropies for the two-dimensional kinetic Ising model with arbitrary single-spin dynamics in the thermodynamic limit are derived as functions of ensemble averages of observables and spin-flip probabilities. Cluster Monte Carlo algorithms with different dynamics from the single-spin dynamics are thus applicable to estimate the transfer entropies. By means of Monte Carlo simulations with the Wolff algorithm, we calculate the information flows in the Ising model with the Metropolis dynamics and the Glauber dynamics, respectively. We find that not only the global Rényi transfer entropy, but also the pairwise Rényi transfer entropy, peaks in the disorder phase.

A hybrid multi-objective imperialist competitive algorithm and Monte Carlo method for robust safety design of a rail vehicle

NASA Astrophysics Data System (ADS)

Nejlaoui, Mohamed; Houidi, Ajmi; Affi, Zouhaier; Romdhane, Lotfi

2017-10-01

This paper deals with the robust safety design optimization of a rail vehicle system moving in short radius curved tracks. A combined multi-objective imperialist competitive algorithm and Monte Carlo method is developed and used for the robust multi-objective optimization of the rail vehicle system. This robust optimization of rail vehicle safety considers simultaneously the derailment angle and its standard deviation where the design parameters uncertainties are considered. The obtained results showed that the robust design reduces significantly the sensitivity of the rail vehicle safety to the design parameters uncertainties compared to the determinist one and to the literature results.

Tracking control of concentration profiles in a fed-batch bioreactor using a linear algebra methodology.

PubMed

Rómoli, Santiago; Serrano, Mario Emanuel; Ortiz, Oscar Alberto; Vega, Jorge Rubén; Eduardo Scaglia, Gustavo Juan

2015-07-01

Based on a linear algebra approach, this paper aims at developing a novel control law able to track reference profiles that were previously-determined in the literature. A main advantage of the proposed strategy is that the control actions are obtained by solving a system of linear equations. The optimal controller parameters are selected through Monte Carlo Randomized Algorithm in order to minimize a proposed cost index. The controller performance is evaluated through several tests, and compared with other controller reported in the literature. Finally, a Monte Carlo Randomized Algorithm is conducted to assess the performance of the proposed controller. Copyright © 2015 ISA. Published by Elsevier Ltd. All rights reserved.

Improved Monte Carlo Scheme for Efficient Particle Transfer in Heterogeneous Systems in the Grand Canonical Ensemble: Application to Vapor-Liquid Nucleation.

PubMed

Loeffler, Troy D; Sepehri, Aliasghar; Chen, Bin

2015-09-08

Reformulation of existing Monte Carlo algorithms used in the study of grand canonical systems has yielded massive improvements in efficiency. Here we present an energy biasing scheme designed to address targeting issues encountered in particle swap moves using sophisticated algorithms such as the Aggregation-Volume-Bias and Unbonding-Bonding methods. Specifically, this energy biasing scheme allows a particle to be inserted to (or removed from) a region that is more acceptable. As a result, this new method showed a several-fold increase in insertion/removal efficiency in addition to an accelerated rate of convergence for the thermodynamic properties of the system.

New estimation architecture for multisensor data fusion

NASA Astrophysics Data System (ADS)

Covino, Joseph M.; Griffiths, Barry E.

1991-07-01

This paper describes a novel method of hierarchical asynchronous distributed filtering called the Net Information Approach (NIA). The NIA is a Kalman-filter-based estimation scheme for spatially distributed sensors which must retain their local optimality yet require a nearly optimal global estimate. The key idea of the NIA is that each local sensor-dedicated filter tells the global filter 'what I've learned since the last local-to-global transmission,' whereas in other estimation architectures the local-to-global transmission consists of 'what I think now.' An algorithm based on this idea has been demonstrated on a small-scale target-tracking problem with many encouraging results. Feasibility of this approach was demonstrated by comparing NIA performance to an optimal centralized Kalman filter (lower bound) via Monte Carlo simulations.

Poster — Thur Eve — 14: Improving Tissue Segmentation for Monte Carlo Dose Calculation using DECT

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

Di Salvio, A.; Bedwani, S.; Carrier, J-F.

2014-08-15

Purpose: To improve Monte Carlo dose calculation accuracy through a new tissue segmentation technique with dual energy CT (DECT). Methods: Electron density (ED) and effective atomic number (EAN) can be extracted directly from DECT data with a stoichiometric calibration method. Images are acquired with Monte Carlo CT projections using the user code egs-cbct and reconstructed using an FDK backprojection algorithm. Calibration is performed using projections of a numerical RMI phantom. A weighted parameter algorithm then uses both EAN and ED to assign materials to voxels from DECT simulated images. This new method is compared to a standard tissue characterization frommore » single energy CT (SECT) data using a segmented calibrated Hounsfield unit (HU) to ED curve. Both methods are compared to the reference numerical head phantom. Monte Carlo simulations on uniform phantoms of different tissues using dosxyz-nrc show discrepancies in depth-dose distributions. Results: Both SECT and DECT segmentation methods show similar performance assigning soft tissues. Performance is however improved with DECT in regions with higher density, such as bones, where it assigns materials correctly 8% more often than segmentation with SECT, considering the same set of tissues and simulated clinical CT images, i.e. including noise and reconstruction artifacts. Furthermore, Monte Carlo results indicate that kV photon beam depth-dose distributions can double between two tissues of density higher than muscle. Conclusions: A direct acquisition of ED and the added information of EAN with DECT data improves tissue segmentation and increases the accuracy of Monte Carlo dose calculation in kV photon beams.« less

Performances of the New Real Time Tsunami Detection Algorithm applied to tide gauges data

NASA Astrophysics Data System (ADS)

Chierici, F.; Embriaco, D.; Morucci, S.

2017-12-01

Real-time tsunami detection algorithms play a key role in any Tsunami Early Warning System. We have developed a new algorithm for tsunami detection (TDA) based on the real-time tide removal and real-time band-pass filtering of seabed pressure time series acquired by Bottom Pressure Recorders. The TDA algorithm greatly increases the tsunami detection probability, shortens the detection delay and enhances detection reliability with respect to the most widely used tsunami detection algorithm, while containing the computational cost. The algorithm is designed to be used also in autonomous early warning systems with a set of input parameters and procedures which can be reconfigured in real time. We have also developed a methodology based on Monte Carlo simulations to test the tsunami detection algorithms. The algorithm performance is estimated by defining and evaluating statistical parameters, namely the detection probability, the detection delay, which are functions of the tsunami amplitude and wavelength, and the occurring rate of false alarms. In this work we present the performance of the TDA algorithm applied to tide gauge data. We have adapted the new tsunami detection algorithm and the Monte Carlo test methodology to tide gauges. Sea level data acquired by coastal tide gauges in different locations and environmental conditions have been used in order to consider real working scenarios in the test. We also present an application of the algorithm to the tsunami event generated by Tohoku earthquake on March 11th 2011, using data recorded by several tide gauges scattered all over the Pacific area.

Assessing the performance of a covert automatic target recognition algorithm

NASA Astrophysics Data System (ADS)

Ehrman, Lisa M.; Lanterman, Aaron D.

2005-05-01

Passive radar systems exploit illuminators of opportunity, such as TV and FM radio, to illuminate potential targets. Doing so allows them to operate covertly and inexpensively. Our research seeks to enhance passive radar systems by adding automatic target recognition (ATR) capabilities. In previous papers we proposed conducting ATR by comparing the radar cross section (RCS) of aircraft detected by a passive radar system to the precomputed RCS of aircraft in the target class. To effectively model the low-frequency setting, the comparison is made via a Rician likelihood model. Monte Carlo simulations indicate that the approach is viable. This paper builds on that work by developing a method for quickly assessing the potential performance of the ATR algorithm without using exhaustive Monte Carlo trials. This method exploits the relation between the probability of error in a binary hypothesis test under the Bayesian framework to the Chernoff information. Since the data are well-modeled as Rician, we begin by deriving a closed-form approximation for the Chernoff information between two Rician densities. This leads to an approximation for the probability of error in the classification algorithm that is a function of the number of available measurements. We conclude with an application that would be particularly cumbersome to accomplish via Monte Carlo trials, but that can be quickly addressed using the Chernoff information approach. This application evaluates the length of time that an aircraft must be tracked before the probability of error in the ATR algorithm drops below a desired threshold.

Subtle Monte Carlo Updates in Dense Molecular Systems.

PubMed

Bottaro, Sandro; Boomsma, Wouter; E Johansson, Kristoffer; Andreetta, Christian; Hamelryck, Thomas; Ferkinghoff-Borg, Jesper

2012-02-14

Although Markov chain Monte Carlo (MC) simulation is a potentially powerful approach for exploring conformational space, it has been unable to compete with molecular dynamics (MD) in the analysis of high density structural states, such as the native state of globular proteins. Here, we introduce a kinetic algorithm, CRISP, that greatly enhances the sampling efficiency in all-atom MC simulations of dense systems. The algorithm is based on an exact analytical solution to the classic chain-closure problem, making it possible to express the interdependencies among degrees of freedom in the molecule as correlations in a multivariate Gaussian distribution. We demonstrate that our method reproduces structural variation in proteins with greater efficiency than current state-of-the-art Monte Carlo methods and has real-time simulation performance on par with molecular dynamics simulations. The presented results suggest our method as a valuable tool in the study of molecules in atomic detail, offering a potential alternative to molecular dynamics for probing long time-scale conformational transitions.

Monte Carlo simulations in X-ray imaging

NASA Astrophysics Data System (ADS)

Giersch, Jürgen; Durst, Jürgen

2008-06-01

Monte Carlo simulations have become crucial tools in many fields of X-ray imaging. They help to understand the influence of physical effects such as absorption, scattering and fluorescence of photons in different detector materials on image quality parameters. They allow studying new imaging concepts like photon counting, energy weighting or material reconstruction. Additionally, they can be applied to the fields of nuclear medicine to define virtual setups studying new geometries or image reconstruction algorithms. Furthermore, an implementation of the propagation physics of electrons and photons allows studying the behavior of (novel) X-ray generation concepts. This versatility of Monte Carlo simulations is illustrated with some examples done by the Monte Carlo simulation ROSI. An overview of the structure of ROSI is given as an example of a modern, well-proven, object-oriented, parallel computing Monte Carlo simulation for X-ray imaging.

Loading relativistic Maxwell distributions in particle simulations

NASA Astrophysics Data System (ADS)

Zenitani, Seiji

2015-04-01

Numerical algorithms to load relativistic Maxwell distributions in particle-in-cell (PIC) and Monte-Carlo simulations are presented. For stationary relativistic Maxwellian, the inverse transform method and the Sobol algorithm are reviewed. To boost particles to obtain relativistic shifted-Maxwellian, two rejection methods are proposed in a physically transparent manner. Their acceptance efficiencies are ≈50 % for generic cases and 100% for symmetric distributions. They can be combined with arbitrary base algorithms.

Polynomial complexity despite the fermionic sign

NASA Astrophysics Data System (ADS)

Rossi, R.; Prokof'ev, N.; Svistunov, B.; Van Houcke, K.; Werner, F.

2017-04-01

It is commonly believed that in unbiased quantum Monte Carlo approaches to fermionic many-body problems, the infamous sign problem generically implies prohibitively large computational times for obtaining thermodynamic-limit quantities. We point out that for convergent Feynman diagrammatic series evaluated with a recently introduced Monte Carlo algorithm (see Rossi R., arXiv:1612.05184), the computational time increases only polynomially with the inverse error on thermodynamic-limit quantities.

Effective Online Bayesian Phylogenetics via Sequential Monte Carlo with Guided Proposals

PubMed Central

Fourment, Mathieu; Claywell, Brian C; Dinh, Vu; McCoy, Connor; Matsen IV, Frederick A; Darling, Aaron E

2018-01-01

Abstract Modern infectious disease outbreak surveillance produces continuous streams of sequence data which require phylogenetic analysis as data arrives. Current software packages for Bayesian phylogenetic inference are unable to quickly incorporate new sequences as they become available, making them less useful for dynamically unfolding evolutionary stories. This limitation can be addressed by applying a class of Bayesian statistical inference algorithms called sequential Monte Carlo (SMC) to conduct online inference, wherein new data can be continuously incorporated to update the estimate of the posterior probability distribution. In this article, we describe and evaluate several different online phylogenetic sequential Monte Carlo (OPSMC) algorithms. We show that proposing new phylogenies with a density similar to the Bayesian prior suffers from poor performance, and we develop “guided” proposals that better match the proposal density to the posterior. Furthermore, we show that the simplest guided proposals can exhibit pathological behavior in some situations, leading to poor results, and that the situation can be resolved by heating the proposal density. The results demonstrate that relative to the widely used MCMC-based algorithm implemented in MrBayes, the total time required to compute a series of phylogenetic posteriors as sequences arrive can be significantly reduced by the use of OPSMC, without incurring a significant loss in accuracy. PMID:29186587

An evaluation of solution algorithms and numerical approximation methods for modeling an ion exchange process

NASA Astrophysics Data System (ADS)

Bu, Sunyoung; Huang, Jingfang; Boyer, Treavor H.; Miller, Cass T.

2010-07-01

The focus of this work is on the modeling of an ion exchange process that occurs in drinking water treatment applications. The model formulation consists of a two-scale model in which a set of microscale diffusion equations representing ion exchange resin particles that vary in size and age are coupled through a boundary condition with a macroscopic ordinary differential equation (ODE), which represents the concentration of a species in a well-mixed reactor. We introduce a new age-averaged model (AAM) that averages all ion exchange particle ages for a given size particle to avoid the expensive Monte-Carlo simulation associated with previous modeling applications. We discuss two different numerical schemes to approximate both the original Monte-Carlo algorithm and the new AAM for this two-scale problem. The first scheme is based on the finite element formulation in space coupled with an existing backward difference formula-based ODE solver in time. The second scheme uses an integral equation based Krylov deferred correction (KDC) method and a fast elliptic solver (FES) for the resulting elliptic equations. Numerical results are presented to validate the new AAM algorithm, which is also shown to be more computationally efficient than the original Monte-Carlo algorithm. We also demonstrate that the higher order KDC scheme is more efficient than the traditional finite element solution approach and this advantage becomes increasingly important as the desired accuracy of the solution increases. We also discuss issues of smoothness, which affect the efficiency of the KDC-FES approach, and outline additional algorithmic changes that would further improve the efficiency of these developing methods for a wide range of applications.

SU-F-SPS-11: The Dosimetric Comparison of Truebeam 2.0 and Cyberknife M6 Treatment Plans for Brain SRS Treatment

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

Mabhouti, H; Sanli, E; Cebe, M

Purpose: Brain stereotactic radiosurgery involves the use of precisely directed, single session radiation to create a desired radiobiologic response within the brain target with acceptable minimal effects on surrounding structures or tissues. In this study, the dosimetric comparison of Truebeam 2.0 and Cyberknife M6 treatment plans were made. Methods: For Truebeam 2.0 machine, treatment planning were done using 2 full arc VMAT technique with 6 FFF beam on the CT scan of Randophantom simulating the treatment of sterotactic treatments for one brain metastasis. The dose distribution were calculated using Eclipse treatment planning system with Acuros XB algorithm. The treatment planningmore » of the same target were also done for Cyberknife M6 machine with Multiplan treatment planning system using Monte Carlo algorithm. Using the same film batch, the net OD to dose calibration curve was obtained using both machine by delivering 0- 800 cGy. Films were scanned 48 hours after irradiation using an Epson 1000XL flatbed scanner. Dose distribution were measured using EBT3 film dosimeter. The measured and calculated doses were compared. Results: The dose distribution in the target and 2 cm beyond the target edge were calculated on TPSs and measured using EBT3 film. For cyberknife plans, the gamma analysis passing rates between measured and calculated dose distributions were 99.2% and 96.7% for target and peripheral region of target respectively. For Truebeam plans, the gamma analysis passing rates were 99.1% and 95.5% for target and peripheral region of target respectively. Conclusion: Although, target dose distribution calculated accurately by Acuros XB and Monte Carlo algorithms, Monte carlo calculation algorithm predicts dose distribution around the peripheral region of target more accurately than Acuros algorithm.« less

Effective optimization using sample persistence: A case study on quantum annealers and various Monte Carlo optimization methods

NASA Astrophysics Data System (ADS)

Karimi, Hamed; Rosenberg, Gili; Katzgraber, Helmut G.

2017-10-01

We present and apply a general-purpose, multistart algorithm for improving the performance of low-energy samplers used for solving optimization problems. The algorithm iteratively fixes the value of a large portion of the variables to values that have a high probability of being optimal. The resulting problems are smaller and less connected, and samplers tend to give better low-energy samples for these problems. The algorithm is trivially parallelizable since each start in the multistart algorithm is independent, and could be applied to any heuristic solver that can be run multiple times to give a sample. We present results for several classes of hard problems solved using simulated annealing, path-integral quantum Monte Carlo, parallel tempering with isoenergetic cluster moves, and a quantum annealer, and show that the success metrics and the scaling are improved substantially. When combined with this algorithm, the quantum annealer's scaling was substantially improved for native Chimera graph problems. In addition, with this algorithm the scaling of the time to solution of the quantum annealer is comparable to the Hamze-de Freitas-Selby algorithm on the weak-strong cluster problems introduced by Boixo et al. Parallel tempering with isoenergetic cluster moves was able to consistently solve three-dimensional spin glass problems with 8000 variables when combined with our method, whereas without our method it could not solve any.

MC3: Multi-core Markov-chain Monte Carlo code

NASA Astrophysics Data System (ADS)

Cubillos, Patricio; Harrington, Joseph; Lust, Nate; Foster, AJ; Stemm, Madison; Loredo, Tom; Stevenson, Kevin; Campo, Chris; Hardin, Matt; Hardy, Ryan

2016-10-01

MC3 (Multi-core Markov-chain Monte Carlo) is a Bayesian statistics tool that can be executed from the shell prompt or interactively through the Python interpreter with single- or multiple-CPU parallel computing. It offers Markov-chain Monte Carlo (MCMC) posterior-distribution sampling for several algorithms, Levenberg-Marquardt least-squares optimization, and uniform non-informative, Jeffreys non-informative, or Gaussian-informative priors. MC3 can share the same value among multiple parameters and fix the value of parameters to constant values, and offers Gelman-Rubin convergence testing and correlated-noise estimation with time-averaging or wavelet-based likelihood estimation methods.

Accelerate quasi Monte Carlo method for solving systems of linear algebraic equations through shared memory

NASA Astrophysics Data System (ADS)

Lai, Siyan; Xu, Ying; Shao, Bo; Guo, Menghan; Lin, Xiaola

2017-04-01

In this paper we study on Monte Carlo method for solving systems of linear algebraic equations (SLAE) based on shared memory. Former research demostrated that GPU can effectively speed up the computations of this issue. Our purpose is to optimize Monte Carlo method simulation on GPUmemoryachritecture specifically. Random numbers are organized to storein shared memory, which aims to accelerate the parallel algorithm. Bank conflicts can be avoided by our Collaborative Thread Arrays(CTA)scheme. The results of experiments show that the shared memory based strategy can speed up the computaions over than 3X at most.

Limited Memory Block Krylov Subspace Optimization for Computing Dominant Singular Value Decompositions

DTIC Science & Technology

2012-03-22

with performance profiles, Math. Program., 91 (2002), pp. 201–213. [6] P. DRINEAS, R. KANNAN, AND M. W. MAHONEY , Fast Monte Carlo algorithms for matrices...computing invariant subspaces of non-Hermitian matri- ces, Numer. Math., 25 ( 1975 /76), pp. 123–136. [25] , Matrix algorithms Vol. II: Eigensystems

Using Computer-Based "Experiments" in the Analysis of Chemical Reaction Equilibria

ERIC Educational Resources Information Center

Li, Zhao; Corti, David S.

2018-01-01

The application of the Reaction Monte Carlo (RxMC) algorithm to standard textbook problems in chemical reaction equilibria is discussed. The RxMC method is a molecular simulation algorithm for studying the equilibrium properties of reactive systems, and therefore provides the opportunity to develop computer-based "experiments" for the…

“Skin-Core-Skin” Structure of Polymer Crystallization Investigated by Multiscale Simulation

PubMed Central

Ruan, Chunlei

2018-01-01

“Skin-core-skin” structure is a typical crystal morphology in injection products. Previous numerical works have rarely focused on crystal evolution; rather, they have mostly been based on the prediction of temperature distribution or crystallization kinetics. The aim of this work was to achieve the “skin-core-skin” structure and investigate the role of external flow and temperature fields on crystal morphology. Therefore, the multiscale algorithm was extended to the simulation of polymer crystallization in a pipe flow. The multiscale algorithm contains two parts: a collocated finite volume method at the macroscopic level and a morphological Monte Carlo method at the microscopic level. The SIMPLE (semi-implicit method for pressure linked equations) algorithm was used to calculate the polymeric model at the macroscopic level, while the Monte Carlo method with stochastic birth-growth process of spherulites and shish-kebabs was used at the microscopic level. Results show that our algorithm is valid to predict “skin-core-skin” structure, and the initial melt temperature and the maximum velocity of melt at the inlet mainly affects the morphology of shish-kebabs. PMID:29659516

A derivation and scalable implementation of the synchronous parallel kinetic Monte Carlo method for simulating long-time dynamics

NASA Astrophysics Data System (ADS)

Byun, Hye Suk; El-Naggar, Mohamed Y.; Kalia, Rajiv K.; Nakano, Aiichiro; Vashishta, Priya

2017-10-01

Kinetic Monte Carlo (KMC) simulations are used to study long-time dynamics of a wide variety of systems. Unfortunately, the conventional KMC algorithm is not scalable to larger systems, since its time scale is inversely proportional to the simulated system size. A promising approach to resolving this issue is the synchronous parallel KMC (SPKMC) algorithm, which makes the time scale size-independent. This paper introduces a formal derivation of the SPKMC algorithm based on local transition-state and time-dependent Hartree approximations, as well as its scalable parallel implementation based on a dual linked-list cell method. The resulting algorithm has achieved a weak-scaling parallel efficiency of 0.935 on 1024 Intel Xeon processors for simulating biological electron transfer dynamics in a 4.2 billion-heme system, as well as decent strong-scaling parallel efficiency. The parallel code has been used to simulate a lattice of cytochrome complexes on a bacterial-membrane nanowire, and it is broadly applicable to other problems such as computational synthesis of new materials.

A global reaction route mapping-based kinetic Monte Carlo algorithm

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

Mitchell, Izaac; Page, Alister J., E-mail: sirle@chem.nagoya-u.ac.jp, E-mail: alister.page@newcastle.edu.au; Irle, Stephan, E-mail: sirle@chem.nagoya-u.ac.jp, E-mail: alister.page@newcastle.edu.au

2016-07-14

We propose a new on-the-fly kinetic Monte Carlo (KMC) method that is based on exhaustive potential energy surface searching carried out with the global reaction route mapping (GRRM) algorithm. Starting from any given equilibrium state, this GRRM-KMC algorithm performs a one-step GRRM search to identify all surrounding transition states. Intrinsic reaction coordinate pathways are then calculated to identify potential subsequent equilibrium states. Harmonic transition state theory is used to calculate rate constants for all potential pathways, before a standard KMC accept/reject selection is performed. The selected pathway is then used to propagate the system forward in time, which is calculatedmore » on the basis of 1st order kinetics. The GRRM-KMC algorithm is validated here in two challenging contexts: intramolecular proton transfer in malonaldehyde and surface carbon diffusion on an iron nanoparticle. We demonstrate that in both cases the GRRM-KMC method is capable of reproducing the 1st order kinetics observed during independent quantum chemical molecular dynamics simulations using the density-functional tight-binding potential.« less

Numerical heating in Particle-In-Cell simulations with Monte Carlo binary collisions

NASA Astrophysics Data System (ADS)

Alves, E. Paulo; Mori, Warren; Fiuza, Frederico

2017-10-01

The binary Monte Carlo collision (BMCC) algorithm is a robust and popular method to include Coulomb collision effects in Particle-in-Cell (PIC) simulations of plasmas. While a number of works have focused on extending the validity of the model to different physical regimes of temperature and density, little attention has been given to the fundamental coupling between PIC and BMCC algorithms. Here, we show that the coupling between PIC and BMCC algorithms can give rise to (nonphysical) numerical heating of the system, that can be far greater than that observed when these algorithms operate independently. This deleterious numerical heating effect can significantly impact the evolution of the simulated system particularly for long simulation times. In this work, we describe the source of this numerical heating, and derive scaling laws for the numerical heating rates based on the numerical parameters of PIC-BMCC simulations. We compare our theoretical scalings with PIC-BMCC numerical experiments, and discuss strategies to minimize this parasitic effect. This work is supported by DOE FES under FWP 100237 and 100182.

A Bootstrap Metropolis-Hastings Algorithm for Bayesian Analysis of Big Data.

PubMed

Liang, Faming; Kim, Jinsu; Song, Qifan

2016-01-01

Markov chain Monte Carlo (MCMC) methods have proven to be a very powerful tool for analyzing data of complex structures. However, their computer-intensive nature, which typically require a large number of iterations and a complete scan of the full dataset for each iteration, precludes their use for big data analysis. In this paper, we propose the so-called bootstrap Metropolis-Hastings (BMH) algorithm, which provides a general framework for how to tame powerful MCMC methods to be used for big data analysis; that is to replace the full data log-likelihood by a Monte Carlo average of the log-likelihoods that are calculated in parallel from multiple bootstrap samples. The BMH algorithm possesses an embarrassingly parallel structure and avoids repeated scans of the full dataset in iterations, and is thus feasible for big data problems. Compared to the popular divide-and-combine method, BMH can be generally more efficient as it can asymptotically integrate the whole data information into a single simulation run. The BMH algorithm is very flexible. Like the Metropolis-Hastings algorithm, it can serve as a basic building block for developing advanced MCMC algorithms that are feasible for big data problems. This is illustrated in the paper by the tempering BMH algorithm, which can be viewed as a combination of parallel tempering and the BMH algorithm. BMH can also be used for model selection and optimization by combining with reversible jump MCMC and simulated annealing, respectively.

A Bootstrap Metropolis–Hastings Algorithm for Bayesian Analysis of Big Data

PubMed Central

Kim, Jinsu; Song, Qifan

2016-01-01

Markov chain Monte Carlo (MCMC) methods have proven to be a very powerful tool for analyzing data of complex structures. However, their computer-intensive nature, which typically require a large number of iterations and a complete scan of the full dataset for each iteration, precludes their use for big data analysis. In this paper, we propose the so-called bootstrap Metropolis-Hastings (BMH) algorithm, which provides a general framework for how to tame powerful MCMC methods to be used for big data analysis; that is to replace the full data log-likelihood by a Monte Carlo average of the log-likelihoods that are calculated in parallel from multiple bootstrap samples. The BMH algorithm possesses an embarrassingly parallel structure and avoids repeated scans of the full dataset in iterations, and is thus feasible for big data problems. Compared to the popular divide-and-combine method, BMH can be generally more efficient as it can asymptotically integrate the whole data information into a single simulation run. The BMH algorithm is very flexible. Like the Metropolis-Hastings algorithm, it can serve as a basic building block for developing advanced MCMC algorithms that are feasible for big data problems. This is illustrated in the paper by the tempering BMH algorithm, which can be viewed as a combination of parallel tempering and the BMH algorithm. BMH can also be used for model selection and optimization by combining with reversible jump MCMC and simulated annealing, respectively. PMID:29033469

Efficient Grammar Induction Algorithm with Parse Forests from Real Corpora

NASA Astrophysics Data System (ADS)

Kurihara, Kenichi; Kameya, Yoshitaka; Sato, Taisuke

The task of inducing grammar structures has received a great deal of attention. The reasons why researchers have studied are different; to use grammar induction as the first stage in building large treebanks or to make up better language models. However, grammar induction has inherent computational complexity. To overcome it, some grammar induction algorithms add new production rules incrementally. They refine the grammar while keeping their computational complexity low. In this paper, we propose a new efficient grammar induction algorithm. Although our algorithm is similar to algorithms which learn a grammar incrementally, our algorithm uses the graphical EM algorithm instead of the Inside-Outside algorithm. We report results of learning experiments in terms of learning speeds. The results show that our algorithm learns a grammar in constant time regardless of the size of the grammar. Since our algorithm decreases syntactic ambiguities in each step, our algorithm reduces required time for learning. This constant-time learning considerably affects learning time for larger grammars. We also reports results of evaluation of criteria to choose nonterminals. Our algorithm refines a grammar based on a nonterminal in each step. Since there can be several criteria to decide which nonterminal is the best, we evaluate them by learning experiments.

Efficient model learning methods for actor-critic control.

PubMed

Grondman, Ivo; Vaandrager, Maarten; Buşoniu, Lucian; Babuska, Robert; Schuitema, Erik

2012-06-01

We propose two new actor-critic algorithms for reinforcement learning. Both algorithms use local linear regression (LLR) to learn approximations of the functions involved. A crucial feature of the algorithms is that they also learn a process model, and this, in combination with LLR, provides an efficient policy update for faster learning. The first algorithm uses a novel model-based update rule for the actor parameters. The second algorithm does not use an explicit actor but learns a reference model which represents a desired behavior, from which desired control actions can be calculated using the inverse of the learned process model. The two novel methods and a standard actor-critic algorithm are applied to the pendulum swing-up problem, in which the novel methods achieve faster learning than the standard algorithm.

TH-E-BRE-07: Development of Dose Calculation Error Predictors for a Widely Implemented Clinical Algorithm

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

Egan, A; Laub, W

2014-06-15

Purpose: Several shortcomings of the current implementation of the analytic anisotropic algorithm (AAA) may lead to dose calculation errors in highly modulated treatments delivered to highly heterogeneous geometries. Here we introduce a set of dosimetric error predictors that can be applied to a clinical treatment plan and patient geometry in order to identify high risk plans. Once a problematic plan is identified, the treatment can be recalculated with more accurate algorithm in order to better assess its viability. Methods: Here we focus on three distinct sources dosimetric error in the AAA algorithm. First, due to a combination of discrepancies inmore » smallfield beam modeling as well as volume averaging effects, dose calculated through small MLC apertures can be underestimated, while that behind small MLC blocks can overestimated. Second, due the rectilinear scaling of the Monte Carlo generated pencil beam kernel, energy is not properly transported through heterogeneities near, but not impeding, the central axis of the beamlet. And third, AAA overestimates dose in regions very low density (< 0.2 g/cm{sup 3}). We have developed an algorithm to detect the location and magnitude of each scenario within the patient geometry, namely the field-size index (FSI), the heterogeneous scatter index (HSI), and the lowdensity index (LDI) respectively. Results: Error indices successfully identify deviations between AAA and Monte Carlo dose distributions in simple phantom geometries. Algorithms are currently implemented in the MATLAB computing environment and are able to run on a typical RapidArc head and neck geometry in less than an hour. Conclusion: Because these error indices successfully identify each type of error in contrived cases, with sufficient benchmarking, this method can be developed into a clinical tool that may be able to help estimate AAA dose calculation errors and when it might be advisable to use Monte Carlo calculations.« less

Identification of moisture content in tobacco plant leaves using outlier sample eliminating algorithms and hyperspectral data.

PubMed

Sun, Jun; Zhou, Xin; Wu, Xiaohong; Zhang, Xiaodong; Li, Qinglin

2016-02-26

Fast identification of moisture content in tobacco plant leaves plays a key role in the tobacco cultivation industry and benefits the management of tobacco plant in the farm. In order to identify moisture content of tobacco plant leaves in a fast and nondestructive way, a method involving Mahalanobis distance coupled with Monte Carlo cross validation(MD-MCCV) was proposed to eliminate outlier sample in this study. The hyperspectral data of 200 tobacco plant leaf samples of 20 moisture gradients were obtained using FieldSpc(®) 3 spectrometer. Savitzky-Golay smoothing(SG), roughness penalty smoothing(RPS), kernel smoothing(KS) and median smoothing(MS) were used to preprocess the raw spectra. In addition, Mahalanobis distance(MD), Monte Carlo cross validation(MCCV) and Mahalanobis distance coupled to Monte Carlo cross validation(MD-MCCV) were applied to select the outlier sample of the raw spectrum and four smoothing preprocessing spectra. Successive projections algorithm (SPA) was used to extract the most influential wavelengths. Multiple Linear Regression (MLR) was applied to build the prediction models based on preprocessed spectra feature in characteristic wavelengths. The results showed that the preferably four prediction model were MD-MCCV-SG (Rp(2) = 0.8401 and RMSEP = 0.1355), MD-MCCV-RPS (Rp(2) = 0.8030 and RMSEP = 0.1274), MD-MCCV-KS (Rp(2) = 0.8117 and RMSEP = 0.1433), MD-MCCV-MS (Rp(2) = 0.9132 and RMSEP = 0.1162). MD-MCCV algorithm performed best among MD algorithm, MCCV algorithm and the method without sample pretreatment algorithm in the eliminating outlier sample from 20 different moisture gradients of tobacco plant leaves and MD-MCCV can be used to eliminate outlier sample in the spectral preprocessing. Copyright © 2016 Elsevier Inc. All rights reserved.

Monte Carlo algorithms for Brownian phylogenetic models.

PubMed

Horvilleur, Benjamin; Lartillot, Nicolas

2014-11-01

Brownian models have been introduced in phylogenetics for describing variation in substitution rates through time, with applications to molecular dating or to the comparative analysis of variation in substitution patterns among lineages. Thus far, however, the Monte Carlo implementations of these models have relied on crude approximations, in which the Brownian process is sampled only at the internal nodes of the phylogeny or at the midpoints along each branch, and the unknown trajectory between these sampled points is summarized by simple branchwise average substitution rates. A more accurate Monte Carlo approach is introduced, explicitly sampling a fine-grained discretization of the trajectory of the (potentially multivariate) Brownian process along the phylogeny. Generic Monte Carlo resampling algorithms are proposed for updating the Brownian paths along and across branches. Specific computational strategies are developed for efficient integration of the finite-time substitution probabilities across branches induced by the Brownian trajectory. The mixing properties and the computational complexity of the resulting Markov chain Monte Carlo sampler scale reasonably with the discretization level, allowing practical applications with up to a few hundred discretization points along the entire depth of the tree. The method can be generalized to other Markovian stochastic processes, making it possible to implement a wide range of time-dependent substitution models with well-controlled computational precision. The program is freely available at www.phylobayes.org. © The Author 2014. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

International Conference on Artificial Immune Systems (1st) ICARIS 2002, held on 9, 10, and 11 September 2002

DTIC Science & Technology

2002-03-07

Michalewicz, Eds., Evolutionary Computation 1: Basic Algorithms and Operators, Institute of Physics, Bristol (UK), 2000. [3] David A. Van Veldhuizen ...2000. [4] Carlos A. Coello Coello, David A. Van Veldhuizen , and Gary B. Lamont, Evolutionary Algorithms for Solving Multi-Objective Problems, Kluwer...Academic Publishers, 233 Spring St., New York, NY 10013, 2002. [5] David A. Van Veldhuizen , Multiobjective Evolution- ary Algorithms: Classifications

Improving multivariate Horner schemes with Monte Carlo tree search

NASA Astrophysics Data System (ADS)

Kuipers, J.; Plaat, A.; Vermaseren, J. A. M.; van den Herik, H. J.

2013-11-01

Optimizing the cost of evaluating a polynomial is a classic problem in computer science. For polynomials in one variable, Horner's method provides a scheme for producing a computationally efficient form. For multivariate polynomials it is possible to generalize Horner's method, but this leaves freedom in the order of the variables. Traditionally, greedy schemes like most-occurring variable first are used. This simple textbook algorithm has given remarkably efficient results. Finding better algorithms has proved difficult. In trying to improve upon the greedy scheme we have implemented Monte Carlo tree search, a recent search method from the field of artificial intelligence. This results in better Horner schemes and reduces the cost of evaluating polynomials, sometimes by factors up to two.

Collective translational and rotational Monte Carlo cluster move for general pairwise interaction

NASA Astrophysics Data System (ADS)

Růžička, Štěpán; Allen, Michael P.

2014-09-01

Virtual move Monte Carlo is a cluster algorithm which was originally developed for strongly attractive colloidal, molecular, or atomistic systems in order to both approximate the collective dynamics and avoid sampling of unphysical kinetic traps. In this paper, we present the algorithm in the form, which selects the moving cluster through a wider class of virtual states and which is applicable to general pairwise interactions, including hard-core repulsion. The newly proposed way of selecting the cluster increases the acceptance probability by up to several orders of magnitude, especially for rotational moves. The results have their applications in simulations of systems interacting via anisotropic potentials both to enhance the sampling of the phase space and to approximate the dynamics.

Improvement of Simulation Method in Validation of Software of the Coordinate Measuring Systems

NASA Astrophysics Data System (ADS)

Nieciąg, Halina

2015-10-01

Software is used in order to accomplish various tasks at each stage of the functioning of modern measuring systems. Before metrological confirmation of measuring equipment, the system has to be validated. This paper discusses the method for conducting validation studies of a fragment of software to calculate the values of measurands. Due to the number and nature of the variables affecting the coordinate measurement results and the complex character and multi-dimensionality of measurands, the study used the Monte Carlo method of numerical simulation. The article presents an attempt of possible improvement of results obtained by classic Monte Carlo tools. The algorithm LHS (Latin Hypercube Sampling) was implemented as alternative to the simple sampling schema of classic algorithm.

A smart Monte Carlo procedure for production costing and uncertainty analysis

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

Parker, C.; Stremel, J.

1996-11-01

Electric utilities using chronological production costing models to decide whether to buy or sell power over the next week or next few weeks need to determine potential profits or losses under a number of uncertainties. A large amount of money can be at stake--often $100,000 a day or more--and one party of the sale must always take on the risk. In the case of fixed price ($/MWh) contracts, the seller accepts the risk. In the case of cost plus contracts, the buyer must accept the risk. So, modeling uncertainty and understanding the risk accurately can improve the competitive edge ofmore » the user. This paper investigates an efficient procedure for representing risks and costs from capacity outages. Typically, production costing models use an algorithm based on some form of random number generator to select resources as available or on outage. These algorithms allow experiments to be repeated and gains and losses to be observed in a short time. The authors perform several experiments to examine the capability of three unit outage selection methods and measures their results. Specifically, a brute force Monte Carlo procedure, a Monte Carlo procedure with Latin Hypercube sampling, and a Smart Monte Carlo procedure with cost stratification and directed sampling are examined.« less

Pushing the limits of Monte Carlo simulations for the three-dimensional Ising model

NASA Astrophysics Data System (ADS)

Ferrenberg, Alan M.; Xu, Jiahao; Landau, David P.

2018-04-01

While the three-dimensional Ising model has defied analytic solution, various numerical methods like Monte Carlo, Monte Carlo renormalization group, and series expansion have provided precise information about the phase transition. Using Monte Carlo simulation that employs the Wolff cluster flipping algorithm with both 32-bit and 53-bit random number generators and data analysis with histogram reweighting and quadruple precision arithmetic, we have investigated the critical behavior of the simple cubic Ising Model, with lattice sizes ranging from 163 to 10243. By analyzing data with cross correlations between various thermodynamic quantities obtained from the same data pool, e.g., logarithmic derivatives of magnetization and derivatives of magnetization cumulants, we have obtained the critical inverse temperature Kc=0.221 654 626 (5 ) and the critical exponent of the correlation length ν =0.629 912 (86 ) with precision that exceeds all previous Monte Carlo estimates.

Markov Chain Monte Carlo Methods for Bayesian Data Analysis in Astronomy

NASA Astrophysics Data System (ADS)

Sharma, Sanjib

2017-08-01

Markov Chain Monte Carlo based Bayesian data analysis has now become the method of choice for analyzing and interpreting data in almost all disciplines of science. In astronomy, over the last decade, we have also seen a steady increase in the number of papers that employ Monte Carlo based Bayesian analysis. New, efficient Monte Carlo based methods are continuously being developed and explored. In this review, we first explain the basics of Bayesian theory and discuss how to set up data analysis problems within this framework. Next, we provide an overview of various Monte Carlo based methods for performing Bayesian data analysis. Finally, we discuss advanced ideas that enable us to tackle complex problems and thus hold great promise for the future. We also distribute downloadable computer software (available at https://github.com/sanjibs/bmcmc/ ) that implements some of the algorithms and examples discussed here.

QMCPACK: an open source ab initio quantum Monte Carlo package for the electronic structure of atoms, molecules and solids

NASA Astrophysics Data System (ADS)

Kim, Jeongnim; Baczewski, Andrew D.; Beaudet, Todd D.; Benali, Anouar; Chandler Bennett, M.; Berrill, Mark A.; Blunt, Nick S.; Josué Landinez Borda, Edgar; Casula, Michele; Ceperley, David M.; Chiesa, Simone; Clark, Bryan K.; Clay, Raymond C., III; Delaney, Kris T.; Dewing, Mark; Esler, Kenneth P.; Hao, Hongxia; Heinonen, Olle; Kent, Paul R. C.; Krogel, Jaron T.; Kylänpää, Ilkka; Li, Ying Wai; Lopez, M. Graham; Luo, Ye; Malone, Fionn D.; Martin, Richard M.; Mathuriya, Amrita; McMinis, Jeremy; Melton, Cody A.; Mitas, Lubos; Morales, Miguel A.; Neuscamman, Eric; Parker, William D.; Pineda Flores, Sergio D.; Romero, Nichols A.; Rubenstein, Brenda M.; Shea, Jacqueline A. R.; Shin, Hyeondeok; Shulenburger, Luke; Tillack, Andreas F.; Townsend, Joshua P.; Tubman, Norm M.; Van Der Goetz, Brett; Vincent, Jordan E.; ChangMo Yang, D.; Yang, Yubo; Zhang, Shuai; Zhao, Luning

2018-05-01

QMCPACK is an open source quantum Monte Carlo package for ab initio electronic structure calculations. It supports calculations of metallic and insulating solids, molecules, atoms, and some model Hamiltonians. Implemented real space quantum Monte Carlo algorithms include variational, diffusion, and reptation Monte Carlo. QMCPACK uses Slater–Jastrow type trial wavefunctions in conjunction with a sophisticated optimizer capable of optimizing tens of thousands of parameters. The orbital space auxiliary-field quantum Monte Carlo method is also implemented, enabling cross validation between different highly accurate methods. The code is specifically optimized for calculations with large numbers of electrons on the latest high performance computing architectures, including multicore central processing unit and graphical processing unit systems. We detail the program’s capabilities, outline its structure, and give examples of its use in current research calculations. The package is available at http://qmcpack.org.

QMCPACK: an open source ab initio quantum Monte Carlo package for the electronic structure of atoms, molecules and solids.

PubMed

Kim, Jeongnim; Baczewski, Andrew T; Beaudet, Todd D; Benali, Anouar; Bennett, M Chandler; Berrill, Mark A; Blunt, Nick S; Borda, Edgar Josué Landinez; Casula, Michele; Ceperley, David M; Chiesa, Simone; Clark, Bryan K; Clay, Raymond C; Delaney, Kris T; Dewing, Mark; Esler, Kenneth P; Hao, Hongxia; Heinonen, Olle; Kent, Paul R C; Krogel, Jaron T; Kylänpää, Ilkka; Li, Ying Wai; Lopez, M Graham; Luo, Ye; Malone, Fionn D; Martin, Richard M; Mathuriya, Amrita; McMinis, Jeremy; Melton, Cody A; Mitas, Lubos; Morales, Miguel A; Neuscamman, Eric; Parker, William D; Pineda Flores, Sergio D; Romero, Nichols A; Rubenstein, Brenda M; Shea, Jacqueline A R; Shin, Hyeondeok; Shulenburger, Luke; Tillack, Andreas F; Townsend, Joshua P; Tubman, Norm M; Van Der Goetz, Brett; Vincent, Jordan E; Yang, D ChangMo; Yang, Yubo; Zhang, Shuai; Zhao, Luning

2018-05-16

QMCPACK is an open source quantum Monte Carlo package for ab initio electronic structure calculations. It supports calculations of metallic and insulating solids, molecules, atoms, and some model Hamiltonians. Implemented real space quantum Monte Carlo algorithms include variational, diffusion, and reptation Monte Carlo. QMCPACK uses Slater-Jastrow type trial wavefunctions in conjunction with a sophisticated optimizer capable of optimizing tens of thousands of parameters. The orbital space auxiliary-field quantum Monte Carlo method is also implemented, enabling cross validation between different highly accurate methods. The code is specifically optimized for calculations with large numbers of electrons on the latest high performance computing architectures, including multicore central processing unit and graphical processing unit systems. We detail the program's capabilities, outline its structure, and give examples of its use in current research calculations. The package is available at http://qmcpack.org.

Rare event simulation in radiation transport

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

Kollman, Craig

1993-10-01

This dissertation studies methods for estimating extremely small probabilities by Monte Carlo simulation. Problems in radiation transport typically involve estimating very rare events or the expected value of a random variable which is with overwhelming probability equal to zero. These problems often have high dimensional state spaces and irregular geometries so that analytic solutions are not possible. Monte Carlo simulation must be used to estimate the radiation dosage being transported to a particular location. If the area is well shielded the probability of any one particular particle getting through is very small. Because of the large number of particles involved,more » even a tiny fraction penetrating the shield may represent an unacceptable level of radiation. It therefore becomes critical to be able to accurately estimate this extremely small probability. Importance sampling is a well known technique for improving the efficiency of rare event calculations. Here, a new set of probabilities is used in the simulation runs. The results are multiple by the likelihood ratio between the true and simulated probabilities so as to keep the estimator unbiased. The variance of the resulting estimator is very sensitive to which new set of transition probabilities are chosen. It is shown that a zero variance estimator does exist, but that its computation requires exact knowledge of the solution. A simple random walk with an associated killing model for the scatter of neutrons is introduced. Large deviation results for optimal importance sampling in random walks are extended to the case where killing is present. An adaptive ``learning`` algorithm for implementing importance sampling is given for more general Markov chain models of neutron scatter. For finite state spaces this algorithm is shown to give with probability one, a sequence of estimates converging exponentially fast to the true solution.« less

TH-E-18A-01: Developments in Monte Carlo Methods for Medical Imaging

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

Badal, A; Zbijewski, W; Bolch, W

Monte Carlo simulation methods are widely used in medical physics research and are starting to be implemented in clinical applications such as radiation therapy planning systems. Monte Carlo simulations offer the capability to accurately estimate quantities of interest that are challenging to measure experimentally while taking into account the realistic anatomy of an individual patient. Traditionally, practical application of Monte Carlo simulation codes in diagnostic imaging was limited by the need for large computational resources or long execution times. However, recent advancements in high-performance computing hardware, combined with a new generation of Monte Carlo simulation algorithms and novel postprocessing methods,more » are allowing for the computation of relevant imaging parameters of interest such as patient organ doses and scatter-to-primaryratios in radiographic projections in just a few seconds using affordable computational resources. Programmable Graphics Processing Units (GPUs), for example, provide a convenient, affordable platform for parallelized Monte Carlo executions that yield simulation times on the order of 10{sup 7} xray/ s. Even with GPU acceleration, however, Monte Carlo simulation times can be prohibitive for routine clinical practice. To reduce simulation times further, variance reduction techniques can be used to alter the probabilistic models underlying the x-ray tracking process, resulting in lower variance in the results without biasing the estimates. Other complementary strategies for further reductions in computation time are denoising of the Monte Carlo estimates and estimating (scoring) the quantity of interest at a sparse set of sampling locations (e.g. at a small number of detector pixels in a scatter simulation) followed by interpolation. Beyond reduction of the computational resources required for performing Monte Carlo simulations in medical imaging, the use of accurate representations of patient anatomy is crucial to the virtual generation of medical images and accurate estimation of radiation dose and other imaging parameters. For this, detailed computational phantoms of the patient anatomy must be utilized and implemented within the radiation transport code. Computational phantoms presently come in one of three format types, and in one of four morphometric categories. Format types include stylized (mathematical equation-based), voxel (segmented CT/MR images), and hybrid (NURBS and polygon mesh surfaces). Morphometric categories include reference (small library of phantoms by age at 50th height/weight percentile), patient-dependent (larger library of phantoms at various combinations of height/weight percentiles), patient-sculpted (phantoms altered to match the patient's unique outer body contour), and finally, patient-specific (an exact representation of the patient with respect to both body contour and internal anatomy). The existence and availability of these phantoms represents a very important advance for the simulation of realistic medical imaging applications using Monte Carlo methods. New Monte Carlo simulation codes need to be thoroughly validated before they can be used to perform novel research. Ideally, the validation process would involve comparison of results with those of an experimental measurement, but accurate replication of experimental conditions can be very challenging. It is very common to validate new Monte Carlo simulations by replicating previously published simulation results of similar experiments. This process, however, is commonly problematic due to the lack of sufficient information in the published reports of previous work so as to be able to replicate the simulation in detail. To aid in this process, the AAPM Task Group 195 prepared a report in which six different imaging research experiments commonly performed using Monte Carlo simulations are described and their results provided. The simulation conditions of all six cases are provided in full detail, with all necessary data on material composition, source, geometry, scoring and other parameters provided. The results of these simulations when performed with the four most common publicly available Monte Carlo packages are also provided in tabular form. The Task Group 195 Report will be useful for researchers needing to validate their Monte Carlo work, and for trainees needing to learn Monte Carlo simulation methods. In this symposium we will review the recent advancements in highperformance computing hardware enabling the reduction in computational resources needed for Monte Carlo simulations in medical imaging. We will review variance reduction techniques commonly applied in Monte Carlo simulations of medical imaging systems and present implementation strategies for efficient combination of these techniques with GPU acceleration. Trade-offs involved in Monte Carlo acceleration by means of denoising and “sparse sampling” will be discussed. A method for rapid scatter correction in cone-beam CT (<5 min/scan) will be presented as an illustration of the simulation speeds achievable with optimized Monte Carlo simulations. We will also discuss the development, availability, and capability of the various combinations of computational phantoms for Monte Carlo simulation of medical imaging systems. Finally, we will review some examples of experimental validation of Monte Carlo simulations and will present the AAPM Task Group 195 Report. Learning Objectives: Describe the advances in hardware available for performing Monte Carlo simulations in high performance computing environments. Explain variance reduction, denoising and sparse sampling techniques available for reduction of computational time needed for Monte Carlo simulations of medical imaging. List and compare the computational anthropomorphic phantoms currently available for more accurate assessment of medical imaging parameters in Monte Carlo simulations. Describe experimental methods used for validation of Monte Carlo simulations in medical imaging. Describe the AAPM Task Group 195 Report and its use for validation and teaching of Monte Carlo simulations in medical imaging.« less

A Machine Learning Method for the Prediction of Receptor Activation in the Simulation of Synapses

PubMed Central

Montes, Jesus; Gomez, Elena; Merchán-Pérez, Angel; DeFelipe, Javier; Peña, Jose-Maria

2013-01-01

Chemical synaptic transmission involves the release of a neurotransmitter that diffuses in the extracellular space and interacts with specific receptors located on the postsynaptic membrane. Computer simulation approaches provide fundamental tools for exploring various aspects of the synaptic transmission under different conditions. In particular, Monte Carlo methods can track the stochastic movements of neurotransmitter molecules and their interactions with other discrete molecules, the receptors. However, these methods are computationally expensive, even when used with simplified models, preventing their use in large-scale and multi-scale simulations of complex neuronal systems that may involve large numbers of synaptic connections. We have developed a machine-learning based method that can accurately predict relevant aspects of the behavior of synapses, such as the percentage of open synaptic receptors as a function of time since the release of the neurotransmitter, with considerably lower computational cost compared with the conventional Monte Carlo alternative. The method is designed to learn patterns and general principles from a corpus of previously generated Monte Carlo simulations of synapses covering a wide range of structural and functional characteristics. These patterns are later used as a predictive model of the behavior of synapses under different conditions without the need for additional computationally expensive Monte Carlo simulations. This is performed in five stages: data sampling, fold creation, machine learning, validation and curve fitting. The resulting procedure is accurate, automatic, and it is general enough to predict synapse behavior under experimental conditions that are different to the ones it has been trained on. Since our method efficiently reproduces the results that can be obtained with Monte Carlo simulations at a considerably lower computational cost, it is suitable for the simulation of high numbers of synapses and it is therefore an excellent tool for multi-scale simulations. PMID:23894367

SemiBoost: boosting for semi-supervised learning.

PubMed

Mallapragada, Pavan Kumar; Jin, Rong; Jain, Anil K; Liu, Yi

2009-11-01

Semi-supervised learning has attracted a significant amount of attention in pattern recognition and machine learning. Most previous studies have focused on designing special algorithms to effectively exploit the unlabeled data in conjunction with labeled data. Our goal is to improve the classification accuracy of any given supervised learning algorithm by using the available unlabeled examples. We call this as the Semi-supervised improvement problem, to distinguish the proposed approach from the existing approaches. We design a metasemi-supervised learning algorithm that wraps around the underlying supervised algorithm and improves its performance using unlabeled data. This problem is particularly important when we need to train a supervised learning algorithm with a limited number of labeled examples and a multitude of unlabeled examples. We present a boosting framework for semi-supervised learning, termed as SemiBoost. The key advantages of the proposed semi-supervised learning approach are: 1) performance improvement of any supervised learning algorithm with a multitude of unlabeled data, 2) efficient computation by the iterative boosting algorithm, and 3) exploiting both manifold and cluster assumption in training classification models. An empirical study on 16 different data sets and text categorization demonstrates that the proposed framework improves the performance of several commonly used supervised learning algorithms, given a large number of unlabeled examples. We also show that the performance of the proposed algorithm, SemiBoost, is comparable to the state-of-the-art semi-supervised learning algorithms.

Using the Metropolis Algorithm to Calculate Thermodynamic Quantities: An Undergraduate Computational Experiment

ERIC Educational Resources Information Center

Beddard, Godfrey S.

2011-01-01

Thermodynamic quantities such as the average energy, heat capacity, and entropy are calculated using a Monte Carlo method based on the Metropolis algorithm. This method is illustrated with reference to the harmonic oscillator but is particularly useful when the partition function cannot be evaluated; an example using a one-dimensional spin system…

Comparison of normal tissue dose calculation methods for epidemiological studies of radiotherapy patients.

PubMed

Mille, Matthew M; Jung, Jae Won; Lee, Choonik; Kuzmin, Gleb A; Lee, Choonsik

2018-06-01

Radiation dosimetry is an essential input for epidemiological studies of radiotherapy patients aimed at quantifying the dose-response relationship of late-term morbidity and mortality. Individualised organ dose must be estimated for all tissues of interest located in-field, near-field, or out-of-field. Whereas conventional measurement approaches are limited to points in water or anthropomorphic phantoms, computational approaches using patient images or human phantoms offer greater flexibility and can provide more detailed three-dimensional dose information. In the current study, we systematically compared four different dose calculation algorithms so that dosimetrists and epidemiologists can better understand the advantages and limitations of the various approaches at their disposal. The four dose calculations algorithms considered were as follows: the (1) Analytical Anisotropic Algorithm (AAA) and (2) Acuros XB algorithm (Acuros XB), as implemented in the Eclipse treatment planning system (TPS); (3) a Monte Carlo radiation transport code, EGSnrc; and (4) an accelerated Monte Carlo code, the x-ray Voxel Monte Carlo (XVMC). The four algorithms were compared in terms of their accuracy and appropriateness in the context of dose reconstruction for epidemiological investigations. Accuracy in peripheral dose was evaluated first by benchmarking the calculated dose profiles against measurements in a homogeneous water phantom. Additional simulations in a heterogeneous cylinder phantom evaluated the performance of the algorithms in the presence of tissue heterogeneity. In general, we found that the algorithms contained within the commercial TPS (AAA and Acuros XB) were fast and accurate in-field or near-field, but not acceptable out-of-field. Therefore, the TPS is best suited for epidemiological studies involving large cohorts and where the organs of interest are located in-field or partially in-field. The EGSnrc and XVMC codes showed excellent agreement with measurements both in-field and out-of-field. The EGSnrc code was the most accurate dosimetry approach, but was too slow to be used for large-scale epidemiological cohorts. The XVMC code showed similar accuracy to EGSnrc, but was significantly faster, and thus epidemiological applications seem feasible, especially when the organs of interest reside far away from the field edge.

The Construction of Semantic Memory: Grammar-Based Representations Learned from Relational Episodic Information

PubMed Central

Battaglia, Francesco P.; Pennartz, Cyriel M. A.

2011-01-01

After acquisition, memories underlie a process of consolidation, making them more resistant to interference and brain injury. Memory consolidation involves systems-level interactions, most importantly between the hippocampus and associated structures, which takes part in the initial encoding of memory, and the neocortex, which supports long-term storage. This dichotomy parallels the contrast between episodic memory (tied to the hippocampal formation), collecting an autobiographical stream of experiences, and semantic memory, a repertoire of facts and statistical regularities about the world, involving the neocortex at large. Experimental evidence points to a gradual transformation of memories, following encoding, from an episodic to a semantic character. This may require an exchange of information between different memory modules during inactive periods. We propose a theory for such interactions and for the formation of semantic memory, in which episodic memory is encoded as relational data. Semantic memory is modeled as a modified stochastic grammar, which learns to parse episodic configurations expressed as an association matrix. The grammar produces tree-like representations of episodes, describing the relationships between its main constituents at multiple levels of categorization, based on its current knowledge of world regularities. These regularities are learned by the grammar from episodic memory information, through an expectation-maximization procedure, analogous to the inside–outside algorithm for stochastic context-free grammars. We propose that a Monte-Carlo sampling version of this algorithm can be mapped on the dynamics of “sleep replay” of previously acquired information in the hippocampus and neocortex. We propose that the model can reproduce several properties of semantic memory such as decontextualization, top-down processing, and creation of schemata. PMID:21887143

Assessing the Performance of a Machine Learning Algorithm in Identifying Bubbles in Dust Emission

NASA Astrophysics Data System (ADS)

Xu, Duo; Offner, Stella S. R.

2017-12-01

Stellar feedback created by radiation and winds from massive stars plays a significant role in both physical and chemical evolution of molecular clouds. This energy and momentum leaves an identifiable signature (“bubbles”) that affects the dynamics and structure of the cloud. Most bubble searches are performed “by eye,” which is usually time-consuming, subjective, and difficult to calibrate. Automatic classifications based on machine learning make it possible to perform systematic, quantifiable, and repeatable searches for bubbles. We employ a previously developed machine learning algorithm, Brut, and quantitatively evaluate its performance in identifying bubbles using synthetic dust observations. We adopt magnetohydrodynamics simulations, which model stellar winds launching within turbulent molecular clouds, as an input to generate synthetic images. We use a publicly available three-dimensional dust continuum Monte Carlo radiative transfer code, HYPERION, to generate synthetic images of bubbles in three Spitzer bands (4.5, 8, and 24 μm). We designate half of our synthetic bubbles as a training set, which we use to train Brut along with citizen-science data from the Milky Way Project (MWP). We then assess Brut’s accuracy using the remaining synthetic observations. We find that Brut’s performance after retraining increases significantly, and it is able to identify yellow bubbles, which are likely associated with B-type stars. Brut continues to perform well on previously identified high-score bubbles, and over 10% of the MWP bubbles are reclassified as high-confidence bubbles, which were previously marginal or ambiguous detections in the MWP data. We also investigate the influence of the size of the training set, dust model, evolutionary stage, and background noise on bubble identification.

Predicting patchy particle crystals: variable box shape simulations and evolutionary algorithms.

PubMed

Bianchi, Emanuela; Doppelbauer, Günther; Filion, Laura; Dijkstra, Marjolein; Kahl, Gerhard

2012-06-07

We consider several patchy particle models that have been proposed in literature and we investigate their candidate crystal structures in a systematic way. We compare two different algorithms for predicting crystal structures: (i) an approach based on Monte Carlo simulations in the isobaric-isothermal ensemble and (ii) an optimization technique based on ideas of evolutionary algorithms. We show that the two methods are equally successful and provide consistent results on crystalline phases of patchy particle systems.

Small-Noise Analysis and Symmetrization of Implicit Monte Carlo Samplers

DOE PAGES

Goodman, Jonathan; Lin, Kevin K.; Morzfeld, Matthias

2015-07-06

Implicit samplers are algorithms for producing independent, weighted samples from multivariate probability distributions. These are often applied in Bayesian data assimilation algorithms. We use Laplace asymptotic expansions to analyze two implicit samplers in the small noise regime. Our analysis suggests a symmetrization of the algorithms that leads to improved implicit sampling schemes at a relatively small additional cost. Here, computational experiments confirm the theory and show that symmetrization is effective for small noise sampling problems.

Calculated X-ray Intensities Using Monte Carlo Algorithms: A Comparison to Experimental EPMA Data

NASA Technical Reports Server (NTRS)

Carpenter, P. K.

2005-01-01

Monte Carlo (MC) modeling has been used extensively to simulate electron scattering and x-ray emission from complex geometries. Here are presented comparisons between MC results and experimental electron-probe microanalysis (EPMA) measurements as well as phi(rhoz) correction algorithms. Experimental EPMA measurements made on NIST SRM 481 (AgAu) and 482 (CuAu) alloys, at a range of accelerating potential and instrument take-off angles, represent a formal microanalysis data set that has been widely used to develop phi(rhoz) correction algorithms. X-ray intensity data produced by MC simulations represents an independent test of both experimental and phi(rhoz) correction algorithms. The alpha-factor method has previously been used to evaluate systematic errors in the analysis of semiconductor and silicate minerals, and is used here to compare the accuracy of experimental and MC-calculated x-ray data. X-ray intensities calculated by MC are used to generate a-factors using the certificated compositions in the CuAu binary relative to pure Cu and Au standards. MC simulations are obtained using the NIST, WinCasino, and WinXray algorithms; derived x-ray intensities have a built-in atomic number correction, and are further corrected for absorption and characteristic fluorescence using the PAP phi(rhoz) correction algorithm. The Penelope code additionally simulates both characteristic and continuum x-ray fluorescence and thus requires no further correction for use in calculating alpha-factors.

High-Throughput Characterization of Porous Materials Using Graphics Processing Units

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

Kim, Jihan; Martin, Richard L.; Rübel, Oliver

We have developed a high-throughput graphics processing units (GPU) code that can characterize a large database of crystalline porous materials. In our algorithm, the GPU is utilized to accelerate energy grid calculations where the grid values represent interactions (i.e., Lennard-Jones + Coulomb potentials) between gas molecules (i.e., CHmore » $$_{4}$$ and CO$$_{2}$$) and material's framework atoms. Using a parallel flood fill CPU algorithm, inaccessible regions inside the framework structures are identified and blocked based on their energy profiles. Finally, we compute the Henry coefficients and heats of adsorption through statistical Widom insertion Monte Carlo moves in the domain restricted to the accessible space. The code offers significant speedup over a single core CPU code and allows us to characterize a set of porous materials at least an order of magnitude larger than ones considered in earlier studies. For structures selected from such a prescreening algorithm, full adsorption isotherms can be calculated by conducting multiple grand canonical Monte Carlo simulations concurrently within the GPU.« less

Searching for efficient Markov chain Monte Carlo proposal kernels

PubMed Central

Yang, Ziheng; Rodríguez, Carlos E.

2013-01-01

Markov chain Monte Carlo (MCMC) or the Metropolis–Hastings algorithm is a simulation algorithm that has made modern Bayesian statistical inference possible. Nevertheless, the efficiency of different Metropolis–Hastings proposal kernels has rarely been studied except for the Gaussian proposal. Here we propose a unique class of Bactrian kernels, which avoid proposing values that are very close to the current value, and compare their efficiency with a number of proposals for simulating different target distributions, with efficiency measured by the asymptotic variance of a parameter estimate. The uniform kernel is found to be more efficient than the Gaussian kernel, whereas the Bactrian kernel is even better. When optimal scales are used for both, the Bactrian kernel is at least 50% more efficient than the Gaussian. Implementation in a Bayesian program for molecular clock dating confirms the general applicability of our results to generic MCMC algorithms. Our results refute a previous claim that all proposals had nearly identical performance and will prompt further research into efficient MCMC proposals. PMID:24218600

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

Serin, E.; Codel, G.; Mabhouti, H.

Purpose: In small field geometries, the electronic equilibrium can be lost, making it challenging for the dose-calculation algorithm to accurately predict the dose, especially in the presence of tissue heterogeneities. In this study, dosimetric accuracy of Monte Carlo (MC) advanced dose calculation and sequential algorithms of Multiplan treatment planning system were investigated for small radiation fields incident on homogeneous and heterogeneous geometries. Methods: Small open fields of fixed cones of Cyberknife M6 unit 100 to 500 mm2 were used for this study. The fields were incident on in house phantom containing lung, air, and bone inhomogeneities and also homogeneous phantom.more » Using the same film batch, the net OD to dose calibration curve was obtained using CK with the 60 mm fixed cone by delivering 0- 800 cGy. Films were scanned 48 hours after irradiation using an Epson 1000XL flatbed scanner. The dosimetric accuracy of MC and sequential algorithms in the presence of the inhomogeneities was compared against EBT3 film dosimetry Results: Open field tests in a homogeneous phantom showed good agreement between two algorithms and film measurement For MC algorithm, the minimum gamma analysis passing rates between measured and calculated dose distributions were 99.7% and 98.3% for homogeneous and inhomogeneous fields in the case of lung and bone respectively. For sequential algorithm, the minimum gamma analysis passing rates were 98.9% and 92.5% for for homogeneous and inhomogeneous fields respectively for used all cone sizes. In the case of the air heterogeneity, the differences were larger for both calculation algorithms. Overall, when compared to measurement, the MC had better agreement than sequential algorithm. Conclusion: The Monte Carlo calculation algorithm in the Multiplan treatment planning system is an improvement over the existing sequential algorithm. Dose discrepancies were observed for in the presence of air inhomogeneities.« less

Noise-enhanced clustering and competitive learning algorithms.

PubMed

Osoba, Osonde; Kosko, Bart

2013-01-01

Noise can provably speed up convergence in many centroid-based clustering algorithms. This includes the popular k-means clustering algorithm. The clustering noise benefit follows from the general noise benefit for the expectation-maximization algorithm because many clustering algorithms are special cases of the expectation-maximization algorithm. Simulations show that noise also speeds up convergence in stochastic unsupervised competitive learning, supervised competitive learning, and differential competitive learning. Copyright © 2012 Elsevier Ltd. All rights reserved.

Q-Learning-Based Adjustable Fixed-Phase Quantum Grover Search Algorithm

NASA Astrophysics Data System (ADS)

Guo, Ying; Shi, Wensha; Wang, Yijun; Hu, Jiankun

2017-02-01

We demonstrate that the rotation phase can be suitably chosen to increase the efficiency of the phase-based quantum search algorithm, leading to a dynamic balance between iterations and success probabilities of the fixed-phase quantum Grover search algorithm with Q-learning for a given number of solutions. In this search algorithm, the proposed Q-learning algorithm, which is a model-free reinforcement learning strategy in essence, is used for performing a matching algorithm based on the fraction of marked items λ and the rotation phase α. After establishing the policy function α = π(λ), we complete the fixed-phase Grover algorithm, where the phase parameter is selected via the learned policy. Simulation results show that the Q-learning-based Grover search algorithm (QLGA) enables fewer iterations and gives birth to higher success probabilities. Compared with the conventional Grover algorithms, it avoids the optimal local situations, thereby enabling success probabilities to approach one.

Excited states from quantum Monte Carlo in the basis of Slater determinants

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

Humeniuk, Alexander; Mitrić, Roland, E-mail: roland.mitric@uni-wuerzburg.de

2014-11-21

Building on the full configuration interaction quantum Monte Carlo (FCIQMC) algorithm introduced recently by Booth et al. [J. Chem. Phys. 131, 054106 (2009)] to compute the ground state of correlated many-electron systems, an extension to the computation of excited states (exFCIQMC) is presented. The Hilbert space is divided into a large part consisting of pure Slater determinants and a much smaller orthogonal part (the size of which is controlled by a cut-off threshold), from which the lowest eigenstates can be removed efficiently. In this way, the quantum Monte Carlo algorithm is restricted to the orthogonal complement of the lower excitedmore » states and projects out the next highest excited state. Starting from the ground state, higher excited states can be found one after the other. The Schrödinger equation in imaginary time is solved by the same population dynamics as in the ground state algorithm with modified probabilities and matrix elements, for which working formulae are provided. As a proof of principle, the method is applied to lithium hydride in the 3-21G basis set and to the helium dimer in the aug-cc-pVDZ basis set. It is shown to give the correct electronic structure for all bond lengths. Much more testing will be required before the applicability of this method to electron correlation problems of interesting size can be assessed.« less

A fast Monte Carlo EM algorithm for estimation in latent class model analysis with an application to assess diagnostic accuracy for cervical neoplasia in women with AGC

PubMed Central

Kang, Le; Carter, Randy; Darcy, Kathleen; Kauderer, James; Liao, Shu-Yuan

2013-01-01

In this article we use a latent class model (LCM) with prevalence modeled as a function of covariates to assess diagnostic test accuracy in situations where the true disease status is not observed, but observations on three or more conditionally independent diagnostic tests are available. A fast Monte Carlo EM (MCEM) algorithm with binary (disease) diagnostic data is implemented to estimate parameters of interest; namely, sensitivity, specificity, and prevalence of the disease as a function of covariates. To obtain standard errors for confidence interval construction of estimated parameters, the missing information principle is applied to adjust information matrix estimates. We compare the adjusted information matrix based standard error estimates with the bootstrap standard error estimates both obtained using the fast MCEM algorithm through an extensive Monte Carlo study. Simulation demonstrates that the adjusted information matrix approach estimates the standard error similarly with the bootstrap methods under certain scenarios. The bootstrap percentile intervals have satisfactory coverage probabilities. We then apply the LCM analysis to a real data set of 122 subjects from a Gynecologic Oncology Group (GOG) study of significant cervical lesion (S-CL) diagnosis in women with atypical glandular cells of undetermined significance (AGC) to compare the diagnostic accuracy of a histology-based evaluation, a CA-IX biomarker-based test and a human papillomavirus (HPV) DNA test. PMID:24163493

Multiscale Monte Carlo equilibration: Two-color QCD with two fermion flavors

DOE PAGES

Detmold, William; Endres, Michael G.

2016-12-02

In this study, we demonstrate the applicability of a recently proposed multiscale thermalization algorithm to two-color quantum chromodynamics (QCD) with two mass-degenerate fermion flavors. The algorithm involves refining an ensemble of gauge configurations that had been generated using a renormalization group (RG) matched coarse action, thereby producing a fine ensemble that is close to the thermalized distribution of a target fine action; the refined ensemble is subsequently rethermalized using conventional algorithms. Although the generalization of this algorithm from pure Yang-Mills theory to QCD with dynamical fermions is straightforward, we find that in the latter case, the method is susceptible tomore » numerical instabilities during the initial stages of rethermalization when using the hybrid Monte Carlo algorithm. We find that these instabilities arise from large fermion forces in the evolution, which are attributed to an accumulation of spurious near-zero modes of the Dirac operator. We propose a simple strategy for curing this problem, and demonstrate that rapid thermalization--as probed by a variety of gluonic and fermionic operators--is possible with the use of this solution. Also, we study the sensitivity of rethermalization rates to the RG matching of the coarse and fine actions, and identify effective matching conditions based on a variety of measured scales.« less

Modification Of Learning Rate With Lvq Model Improvement In Learning Backpropagation

NASA Astrophysics Data System (ADS)

Tata Hardinata, Jaya; Zarlis, Muhammad; Budhiarti Nababan, Erna; Hartama, Dedy; Sembiring, Rahmat W.

2017-12-01

One type of artificial neural network is a backpropagation, This algorithm trained with the network architecture used during the training as well as providing the correct output to insert a similar but not the same with the architecture in use at training.The selection of appropriate parameters also affects the outcome, value of learning rate is one of the parameters which influence the process of training, Learning rate affects the speed of learning process on the network architecture.If the learning rate is set too large, then the algorithm will become unstable and otherwise the algorithm will converge in a very long period of time.So this study was made to determine the value of learning rate on the backpropagation algorithm. LVQ models of learning rate is one of the models used in the determination of the value of the learning rate of the algorithm LVQ.By modifying this LVQ model to be applied to the backpropagation algorithm. From the experimental results known to modify the learning rate LVQ models were applied to the backpropagation algorithm learning process becomes faster (epoch less).

Discrete range clustering using Monte Carlo methods

NASA Technical Reports Server (NTRS)

Chatterji, G. B.; Sridhar, B.

1993-01-01

For automatic obstacle avoidance guidance during rotorcraft low altitude flight, a reliable model of the nearby environment is needed. Such a model may be constructed by applying surface fitting techniques to the dense range map obtained by active sensing using radars. However, for covertness, passive sensing techniques using electro-optic sensors are desirable. As opposed to the dense range map obtained via active sensing, passive sensing algorithms produce reliable range at sparse locations, and therefore, surface fitting techniques to fill the gaps in the range measurement are not directly applicable. Both for automatic guidance and as a display for aiding the pilot, these discrete ranges need to be grouped into sets which correspond to objects in the nearby environment. The focus of this paper is on using Monte Carlo methods for clustering range points into meaningful groups. One of the aims of the paper is to explore whether simulated annealing methods offer significant advantage over the basic Monte Carlo method for this class of problems. We compare three different approaches and present application results of these algorithms to a laboratory image sequence and a helicopter flight sequence.

Hamiltonian Monte Carlo acceleration using surrogate functions with random bases.

PubMed

Zhang, Cheng; Shahbaba, Babak; Zhao, Hongkai

2017-11-01

For big data analysis, high computational cost for Bayesian methods often limits their applications in practice. In recent years, there have been many attempts to improve computational efficiency of Bayesian inference. Here we propose an efficient and scalable computational technique for a state-of-the-art Markov chain Monte Carlo methods, namely, Hamiltonian Monte Carlo. The key idea is to explore and exploit the structure and regularity in parameter space for the underlying probabilistic model to construct an effective approximation of its geometric properties. To this end, we build a surrogate function to approximate the target distribution using properly chosen random bases and an efficient optimization process. The resulting method provides a flexible, scalable, and efficient sampling algorithm, which converges to the correct target distribution. We show that by choosing the basis functions and optimization process differently, our method can be related to other approaches for the construction of surrogate functions such as generalized additive models or Gaussian process models. Experiments based on simulated and real data show that our approach leads to substantially more efficient sampling algorithms compared to existing state-of-the-art methods.

Estimating source parameters from deformation data, with an application to the March 1997 earthquake swarm off the Izu Peninsula, Japan

NASA Astrophysics Data System (ADS)

Cervelli, P.; Murray, M. H.; Segall, P.; Aoki, Y.; Kato, T.

2001-06-01

We have applied two Monte Carlo optimization techniques, simulated annealing and random cost, to the inversion of deformation data for fault and magma chamber geometry. These techniques involve an element of randomness that permits them to escape local minima and ultimately converge to the global minimum of misfit space. We have tested the Monte Carlo algorithms on two synthetic data sets. We have also compared them to one another in terms of their efficiency and reliability. We have applied the bootstrap method to estimate confidence intervals for the source parameters, including the correlations inherent in the data. Additionally, we present methods that use the information from the bootstrapping procedure to visualize the correlations between the different model parameters. We have applied these techniques to GPS, tilt, and leveling data from the March 1997 earthquake swarm off of the Izu Peninsula, Japan. Using the two Monte Carlo algorithms, we have inferred two sources, a dike and a fault, that fit the deformation data and the patterns of seismicity and that are consistent with the regional stress field.

Event identification for KM3NeT/ARCA

NASA Astrophysics Data System (ADS)

Heid, Thomas; KM3NeT Collaboration

2017-09-01

KM3NeT is a large research infrastructure consisting of a network of deep-sea neutrino telescopes. KM3NeT/ARCA will be the instrument detecting high-energy neutrinos with energies above 100 TeV. This instrument gives a new opportunity to observe the neutrino sky with very high angular resolution to be able to detect neutrino point sources. Furthermore it will be possible to probe the flavour composition of neutrino fluxes, and hence production mechanisms, with so-far unreached precision. Neutrinos produce different event topologies in the detector according to their flavour, interaction channel and deposited energy. Machine-learning algorithms are able to learn features of topologies to discriminate them. In previous analyses only two event types were regarded, namely the shower and track topology. With good timing resolution and precise reconstruction algorithms it is possible to separate into more event types, for example the double bang topology produced by tau neutrinos. The final goal is to distinguish all three neutrino flavors as much as possible. To resolve this issue the KM3NeT collaboration uses deep neural networks trained with Monte Carlo events of all neutrino types. This contribution shows the ability of KM3NeT/ARCA to classify events in more than two neutrino event topologies. Furthermore, the borders between detectable classes are shown, such as the minimum distance the tau has to travel before decaying into a tau neutrino to be detected as double bang event.

Predicting the survival of diabetes using neural network

NASA Astrophysics Data System (ADS)

Mamuda, Mamman; Sathasivam, Saratha

2017-08-01

Data mining techniques at the present time are used in predicting diseases of health care industries. Neural Network is one among the prevailing method in data mining techniques of an intelligent field for predicting diseases in health care industries. This paper presents a study on the prediction of the survival of diabetes diseases using different learning algorithms from the supervised learning algorithms of neural network. Three learning algorithms are considered in this study: (i) The levenberg-marquardt learning algorithm (ii) The Bayesian regulation learning algorithm and (iii) The scaled conjugate gradient learning algorithm. The network is trained using the Pima Indian Diabetes Dataset with the help of MATLAB R2014(a) software. The performance of each algorithm is further discussed through regression analysis. The prediction accuracy of the best algorithm is further computed to validate the accurate prediction

Parallelization and implementation of approximate root isolation for nonlinear system by Monte Carlo

NASA Astrophysics Data System (ADS)

Khosravi, Ebrahim

1998-12-01

This dissertation solves a fundamental problem of isolating the real roots of nonlinear systems of equations by Monte-Carlo that were published by Bush Jones. This algorithm requires only function values and can be applied readily to complicated systems of transcendental functions. The implementation of this sequential algorithm provides scientists with the means to utilize function analysis in mathematics or other fields of science. The algorithm, however, is so computationally intensive that the system is limited to a very small set of variables, and this will make it unfeasible for large systems of equations. Also a computational technique was needed for investigating a metrology of preventing the algorithm structure from converging to the same root along different paths of computation. The research provides techniques for improving the efficiency and correctness of the algorithm. The sequential algorithm for this technique was corrected and a parallel algorithm is presented. This parallel method has been formally analyzed and is compared with other known methods of root isolation. The effectiveness, efficiency, enhanced overall performance of the parallel processing of the program in comparison to sequential processing is discussed. The message passing model was used for this parallel processing, and it is presented and implemented on Intel/860 MIMD architecture. The parallel processing proposed in this research has been implemented in an ongoing high energy physics experiment: this algorithm has been used to track neutrinoes in a super K detector. This experiment is located in Japan, and data can be processed on-line or off-line locally or remotely.

Accurate condensed history Monte Carlo simulation of electron transport. II. Application to ion chamber response simulations.

PubMed

Kawrakow, I

2000-03-01

In this report the condensed history Monte Carlo simulation of electron transport and its application to the calculation of ion chamber response is discussed. It is shown that the strong step-size dependencies and lack of convergence to the correct answer previously observed are the combined effect of the following artifacts caused by the EGS4/PRESTA implementation of the condensed history technique: dose underprediction due to PRESTA'S pathlength correction and lateral correlation algorithm; dose overprediction due to the boundary crossing algorithm; dose overprediction due to the breakdown of the fictitious cross section method for sampling distances between discrete interaction and the inaccurate evaluation of energy-dependent quantities. These artifacts are now understood quantitatively and analytical expressions for their effect are given.

Parameters estimation for reactive transport: A way to test the validity of a reactive model

NASA Astrophysics Data System (ADS)

Aggarwal, Mohit; Cheikh Anta Ndiaye, Mame; Carrayrou, Jérôme

The chemical parameters used in reactive transport models are not known accurately due to the complexity and the heterogeneous conditions of a real domain. We will present an efficient algorithm in order to estimate the chemical parameters using Monte-Carlo method. Monte-Carlo methods are very robust for the optimisation of the highly non-linear mathematical model describing reactive transport. Reactive transport of tributyltin (TBT) through natural quartz sand at seven different pHs is taken as the test case. Our algorithm will be used to estimate the chemical parameters of the sorption of TBT onto the natural quartz sand. By testing and comparing three models of surface complexation, we show that the proposed adsorption model cannot explain the experimental data.

Analysis of estimation algorithms for CDTI and CAS applications

NASA Technical Reports Server (NTRS)

Goka, T.

1985-01-01

Estimation algorithms for Cockpit Display of Traffic Information (CDTI) and Collision Avoidance System (CAS) applications were analyzed and/or developed. The algorithms are based on actual or projected operational and performance characteristics of an Enhanced TCAS II traffic sensor developed by Bendix and the Federal Aviation Administration. Three algorithm areas are examined and discussed. These are horizontal x and y, range and altitude estimation algorithms. Raw estimation errors are quantified using Monte Carlo simulations developed for each application; the raw errors are then used to infer impacts on the CDTI and CAS applications. Applications of smoothing algorithms to CDTI problems are also discussed briefly. Technical conclusions are summarized based on the analysis of simulation results.

Recommendations for dose calculations of lung cancer treatment plans treated with stereotactic ablative body radiotherapy (SABR)

NASA Astrophysics Data System (ADS)

Devpura, S.; Siddiqui, M. S.; Chen, D.; Liu, D.; Li, H.; Kumar, S.; Gordon, J.; Ajlouni, M.; Movsas, B.; Chetty, I. J.

2014-03-01

The purpose of this study was to systematically evaluate dose distributions computed with 5 different dose algorithms for patients with lung cancers treated using stereotactic ablative body radiotherapy (SABR). Treatment plans for 133 lung cancer patients, initially computed with a 1D-pencil beam (equivalent-path-length, EPL-1D) algorithm, were recalculated with 4 other algorithms commissioned for treatment planning, including 3-D pencil-beam (EPL-3D), anisotropic analytical algorithm (AAA), collapsed cone convolution superposition (CCC), and Monte Carlo (MC). The plan prescription dose was 48 Gy in 4 fractions normalized to the 95% isodose line. Tumors were classified according to location: peripheral tumors surrounded by lung (lung-island, N=39), peripheral tumors attached to the rib-cage or chest wall (lung-wall, N=44), and centrally-located tumors (lung-central, N=50). Relative to the EPL-1D algorithm, PTV D95 and mean dose values computed with the other 4 algorithms were lowest for "lung-island" tumors with smallest field sizes (3-5 cm). On the other hand, the smallest differences were noted for lung-central tumors treated with largest field widths (7-10 cm). Amongst all locations, dose distribution differences were most strongly correlated with tumor size for lung-island tumors. For most cases, convolution/superposition and MC algorithms were in good agreement. Mean lung dose (MLD) values computed with the EPL-1D algorithm were highly correlated with that of the other algorithms (correlation coefficient =0.99). The MLD values were found to be ~10% lower for small lung-island tumors with the model-based (conv/superposition and MC) vs. the correction-based (pencil-beam) algorithms with the model-based algorithms predicting greater low dose spread within the lungs. This study suggests that pencil beam algorithms should be avoided for lung SABR planning. For the most challenging cases, small tumors surrounded entirely by lung tissue (lung-island type), a Monte-Carlo-based algorithm may be warranted.

A Monte Carlo comparison of the recovery of winds near upwind and downwind from the SASS-1 model function by means of the sum of squares algorithm and a maximum likelihood estimator

NASA Technical Reports Server (NTRS)

Pierson, W. J., Jr.

1984-01-01

Backscatter measurements at upwind and crosswind are simulated for five incidence angles by means of the SASS-1 model function. The effects of communication noise and attitude errors are simulated by Monte Carlo methods, and the winds are recovered by both the Sum of Square (SOS) algorithm and a Maximum Likelihood Estimater (MLE). The SOS algorithm is shown to fail for light enough winds at all incidence angles and to fail to show areas of calm because backscatter estimates that were negative or that produced incorrect values of K sub p greater than one were discarded. The MLE performs well for all input backscatter estimates and returns calm when both are negative. The use of the SOS algorithm is shown to have introduced errors in the SASS-1 model function that, in part, cancel out the errors that result from using it, but that also cause disagreement with other data sources such as the AAFE circle flight data at light winds. Implications for future scatterometer systems are given.

Asteroid mass estimation using Markov-chain Monte Carlo

NASA Astrophysics Data System (ADS)

Siltala, Lauri; Granvik, Mikael

2017-11-01

Estimates for asteroid masses are based on their gravitational perturbations on the orbits of other objects such as Mars, spacecraft, or other asteroids and/or their satellites. In the case of asteroid-asteroid perturbations, this leads to an inverse problem in at least 13 dimensions where the aim is to derive the mass of the perturbing asteroid(s) and six orbital elements for both the perturbing asteroid(s) and the test asteroid(s) based on astrometric observations. We have developed and implemented three different mass estimation algorithms utilizing asteroid-asteroid perturbations: the very rough 'marching' approximation, in which the asteroids' orbital elements are not fitted, thereby reducing the problem to a one-dimensional estimation of the mass, an implementation of the Nelder-Mead simplex method, and most significantly, a Markov-chain Monte Carlo (MCMC) approach. We describe each of these algorithms with particular focus on the MCMC algorithm, and present example results using both synthetic and real data. Our results agree with the published mass estimates, but suggest that the published uncertainties may be misleading as a consequence of using linearized mass-estimation methods. Finally, we discuss remaining challenges with the algorithms as well as future plans.

Multi-particle phase space integration with arbitrary set of singularities in CompHEP

NASA Astrophysics Data System (ADS)

Kovalenko, D. N.; Pukhov, A. E.

1997-02-01

We describe an algorithm of multi-particle phase space integration for collision and decay processes realized in CompHEP package version 3.2. In the framework of this algorithm it is possible to regularize an arbitrary set of singularities caused by virtual particle propagators. The algorithm is based on the method of the recursive representation of kinematics and on the multichannel Monte Carlo approach. CompHEP package is available by WWW: http://theory.npi.msu.su/pukhov/comphep.html

An Educational System for Learning Search Algorithms and Automatically Assessing Student Performance

ERIC Educational Resources Information Center

Grivokostopoulou, Foteini; Perikos, Isidoros; Hatzilygeroudis, Ioannis

2017-01-01

In this paper, first we present an educational system that assists students in learning and tutors in teaching search algorithms, an artificial intelligence topic. Learning is achieved through a wide range of learning activities. Algorithm visualizations demonstrate the operational functionality of algorithms according to the principles of active…

Using an improved association rules mining optimization algorithm in web-based mobile-learning system

NASA Astrophysics Data System (ADS)

Huang, Yin; Chen, Jianhua; Xiong, Shaojun

2009-07-01

Mobile-Learning (M-learning) makes many learners get the advantages of both traditional learning and E-learning. Currently, Web-based Mobile-Learning Systems have created many new ways and defined new relationships between educators and learners. Association rule mining is one of the most important fields in data mining and knowledge discovery in databases. Rules explosion is a serious problem which causes great concerns, as conventional mining algorithms often produce too many rules for decision makers to digest. Since Web-based Mobile-Learning System collects vast amounts of student profile data, data mining and knowledge discovery techniques can be applied to find interesting relationships between attributes of learners, assessments, the solution strategies adopted by learners and so on. Therefore ,this paper focus on a new data-mining algorithm, combined with the advantages of genetic algorithm and simulated annealing algorithm , called ARGSA(Association rules based on an improved Genetic Simulated Annealing Algorithm), to mine the association rules. This paper first takes advantage of the Parallel Genetic Algorithm and Simulated Algorithm designed specifically for discovering association rules. Moreover, the analysis and experiment are also made to show the proposed method is superior to the Apriori algorithm in this Mobile-Learning system.

Assessing Continuous Operator Workload With a Hybrid Scaffolded Neuroergonomic Modeling Approach.

PubMed

Borghetti, Brett J; Giametta, Joseph J; Rusnock, Christina F

2017-02-01

We aimed to predict operator workload from neurological data using statistical learning methods to fit neurological-to-state-assessment models. Adaptive systems require real-time mental workload assessment to perform dynamic task allocations or operator augmentation as workload issues arise. Neuroergonomic measures have great potential for informing adaptive systems, and we combine these measures with models of task demand as well as information about critical events and performance to clarify the inherent ambiguity of interpretation. We use machine learning algorithms on electroencephalogram (EEG) input to infer operator workload based upon Improved Performance Research Integration Tool workload model estimates. Cross-participant models predict workload of other participants, statistically distinguishing between 62% of the workload changes. Machine learning models trained from Monte Carlo resampled workload profiles can be used in place of deterministic workload profiles for cross-participant modeling without incurring a significant decrease in machine learning model performance, suggesting that stochastic models can be used when limited training data are available. We employed a novel temporary scaffold of simulation-generated workload profile truth data during the model-fitting process. A continuous workload profile serves as the target to train our statistical machine learning models. Once trained, the workload profile scaffolding is removed and the trained model is used directly on neurophysiological data in future operator state assessments. These modeling techniques demonstrate how to use neuroergonomic methods to develop operator state assessments, which can be employed in adaptive systems.

Monte Carlo Calculations of Polarized Microwave Radiation Emerging from Cloud Structures

NASA Technical Reports Server (NTRS)

Kummerow, Christian; Roberti, Laura

1998-01-01

The last decade has seen tremendous growth in cloud dynamical and microphysical models that are able to simulate storms and storm systems with very high spatial resolution, typically of the order of a few kilometers. The fairly realistic distributions of cloud and hydrometeor properties that these models generate has in turn led to a renewed interest in the three-dimensional microwave radiative transfer modeling needed to understand the effect of cloud and rainfall inhomogeneities upon microwave observations. Monte Carlo methods, and particularly backwards Monte Carlo methods have shown themselves to be very desirable due to the quick convergence of the solutions. Unfortunately, backwards Monte Carlo methods are not well suited to treat polarized radiation. This study reviews the existing Monte Carlo methods and presents a new polarized Monte Carlo radiative transfer code. The code is based on a forward scheme but uses aliasing techniques to keep the computational requirements equivalent to the backwards solution. Radiative transfer computations have been performed using a microphysical-dynamical cloud model and the results are presented together with the algorithm description.

A comparison of back propagation and Generalized Regression Neural Networks performance in neutron spectrometry.

PubMed

Martínez-Blanco, Ma Del Rosario; Ornelas-Vargas, Gerardo; Solís-Sánchez, Luis Octavio; Castañeda-Miranada, Rodrigo; Vega-Carrillo, Héctor René; Celaya-Padilla, José M; Garza-Veloz, Idalia; Martínez-Fierro, Margarita; Ortiz-Rodríguez, José Manuel

2016-11-01

The process of unfolding the neutron energy spectrum has been subject of research for many years. Monte Carlo, iterative methods, the bayesian theory, the principle of maximum entropy are some of the methods used. The drawbacks associated with traditional unfolding procedures have motivated the research of complementary approaches. Back Propagation Neural Networks (BPNN), have been applied with success in neutron spectrometry and dosimetry domains, however, the structure and learning parameters are factors that highly impact in the networks performance. In ANN domain, Generalized Regression Neural Network (GRNN) is one of the simplest neural networks in term of network architecture and learning algorithm. The learning is instantaneous, requiring no time for training. Opposite to BPNN, a GRNN would be formed instantly with just a 1-pass training on the development data. In the network development phase, the only hurdle is to optimize the hyper-parameter, which is known as sigma, governing the smoothness of the network. The aim of this work was to compare the performance of BPNN and GRNN in the solution of the neutron spectrometry problem. From results obtained it can be observed that despite the very similar results, GRNN performs better than BPNN. Copyright © 2016 Elsevier Ltd. All rights reserved.

Multilevel Monte Carlo and improved timestepping methods in atmospheric dispersion modelling

NASA Astrophysics Data System (ADS)

Katsiolides, Grigoris; Müller, Eike H.; Scheichl, Robert; Shardlow, Tony; Giles, Michael B.; Thomson, David J.

2018-02-01

A common way to simulate the transport and spread of pollutants in the atmosphere is via stochastic Lagrangian dispersion models. Mathematically, these models describe turbulent transport processes with stochastic differential equations (SDEs). The computational bottleneck is the Monte Carlo algorithm, which simulates the motion of a large number of model particles in a turbulent velocity field; for each particle, a trajectory is calculated with a numerical timestepping method. Choosing an efficient numerical method is particularly important in operational emergency-response applications, such as tracking radioactive clouds from nuclear accidents or predicting the impact of volcanic ash clouds on international aviation, where accurate and timely predictions are essential. In this paper, we investigate the application of the Multilevel Monte Carlo (MLMC) method to simulate the propagation of particles in a representative one-dimensional dispersion scenario in the atmospheric boundary layer. MLMC can be shown to result in asymptotically superior computational complexity and reduced computational cost when compared to the Standard Monte Carlo (StMC) method, which is currently used in atmospheric dispersion modelling. To reduce the absolute cost of the method also in the non-asymptotic regime, it is equally important to choose the best possible numerical timestepping method on each level. To investigate this, we also compare the standard symplectic Euler method, which is used in many operational models, with two improved timestepping algorithms based on SDE splitting methods.

A Circuit-Based Neural Network with Hybrid Learning of Backpropagation and Random Weight Change Algorithms

PubMed Central

Yang, Changju; Kim, Hyongsuk; Adhikari, Shyam Prasad; Chua, Leon O.

2016-01-01

A hybrid learning method of a software-based backpropagation learning and a hardware-based RWC learning is proposed for the development of circuit-based neural networks. The backpropagation is known as one of the most efficient learning algorithms. A weak point is that its hardware implementation is extremely difficult. The RWC algorithm, which is very easy to implement with respect to its hardware circuits, takes too many iterations for learning. The proposed learning algorithm is a hybrid one of these two. The main learning is performed with a software version of the BP algorithm, firstly, and then, learned weights are transplanted on a hardware version of a neural circuit. At the time of the weight transplantation, a significant amount of output error would occur due to the characteristic difference between the software and the hardware. In the proposed method, such error is reduced via a complementary learning of the RWC algorithm, which is implemented in a simple hardware. The usefulness of the proposed hybrid learning system is verified via simulations upon several classical learning problems. PMID:28025566

Use of Machine Learning Algorithms to Propose a New Methodology to Conduct, Critique and Validate Urban Scale Building Energy Modeling

NASA Astrophysics Data System (ADS)

Pathak, Maharshi

City administrators and real-estate developers have been setting up rather aggressive energy efficiency targets. This, in turn, has led the building science research groups across the globe to focus on urban scale building performance studies and level of abstraction associated with the simulations of the same. The increasing maturity of the stakeholders towards energy efficiency and creating comfortable working environment has led researchers to develop methodologies and tools for addressing the policy driven interventions whether it's urban level energy systems, buildings' operational optimization or retrofit guidelines. Typically, these large-scale simulations are carried out by grouping buildings based on their design similarities i.e. standardization of the buildings. Such an approach does not necessarily lead to potential working inputs which can make decision-making effective. To address this, a novel approach is proposed in the present study. The principle objective of this study is to propose, to define and evaluate the methodology to utilize machine learning algorithms in defining representative building archetypes for the Stock-level Building Energy Modeling (SBEM) which are based on operational parameter database. The study uses "Phoenix- climate" based CBECS-2012 survey microdata for analysis and validation. Using the database, parameter correlations are studied to understand the relation between input parameters and the energy performance. Contrary to precedence, the study establishes that the energy performance is better explained by the non-linear models. The non-linear behavior is explained by advanced learning algorithms. Based on these algorithms, the buildings at study are grouped into meaningful clusters. The cluster "mediod" (statistically the centroid, meaning building that can be represented as the centroid of the cluster) are established statistically to identify the level of abstraction that is acceptable for the whole building energy simulations and post that the retrofit decision-making. Further, the methodology is validated by conducting Monte-Carlo simulations on 13 key input simulation parameters. The sensitivity analysis of these 13 parameters is utilized to identify the optimum retrofits. From the sample analysis, the envelope parameters are found to be more sensitive towards the EUI of the building and thus retrofit packages should also be directed to maximize the energy usage reduction.

Superior Generalization Capability of Hardware-Learing Algorithm Developed for Self-Learning Neuron-MOS Neural Networks

NASA Astrophysics Data System (ADS)

Kondo, Shuhei; Shibata, Tadashi; Ohmi, Tadahiro

1995-02-01

We have investigated the learning performance of the hardware backpropagation (HBP) algorithm, a hardware-oriented learning algorithm developed for the self-learning architecture of neural networks constructed using neuron MOS (metal-oxide-semiconductor) transistors. The solution to finding a mirror symmetry axis in a 4×4 binary pixel array was tested by computer simulation based on the HBP algorithm. Despite the inherent restrictions imposed on the hardware-learning algorithm, HBP exhibits equivalent learning performance to that of the original backpropagation (BP) algorithm when all the pertinent parameters are optimized. Very importantly, we have found that HBP has a superior generalization capability over BP; namely, HBP exhibits higher performance in solving problems that the network has not yet learnt.

SU-E-T-344: Validation and Clinical Experience of Eclipse Electron Monte Carlo Algorithm (EMC)

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

Pokharel, S; Rana, S

2014-06-01

Purpose: The purpose of this study is to validate Eclipse Electron Monte Carlo (Algorithm for routine clinical uses. Methods: The PTW inhomogeneity phantom (T40037) with different combination of heterogeneous slabs has been CT-scanned with Philips Brilliance 16 slice scanner. The phantom contains blocks of Rando Alderson materials mimicking lung, Polystyrene (Tissue), PTFE (Bone) and PMAA. The phantom has 30×30×2.5 cm base plate with 2cm recesses to insert inhomogeneity. The detector systems used in this study are diode, tlds and Gafchromic EBT2 films. The diode and tlds were included in CT scans. The CT sets are transferred to Eclipse treatment planningmore » system. Several plans have been created with Eclipse Monte Carlo (EMC) algorithm 11.0.21. Measurements have been carried out in Varian TrueBeam machine for energy from 6–22mev. Results: The measured and calculated doses agreed very well for tissue like media. The agreement was reasonably okay for the presence of lung inhomogeneity. The point dose agreement was within 3.5% and Gamma passing rate at 3%/3mm was greater than 93% except for 6Mev(85%). The disagreement can reach as high as 10% in the presence of bone inhomogeneity. This is due to eclipse reporting dose to the medium as opposed to the dose to the water as in conventional calculation engines. Conclusion: Care must be taken when using Varian Eclipse EMC algorithm for dose calculation for routine clinical uses. The algorithm dose not report dose to water in which most of the clinical experiences are based on rather it just reports dose to medium directly. In the presence of inhomogeneity such as bone, the dose discrepancy can be as high as 10% or even more depending on the location of normalization point or volume. As Radiation oncology as an empirical science, care must be taken before using EMC reported monitor units for clinical uses.« less

Characterisation of mega-voltage electron pencil beam dose distributions: viability of a measurement-based approach.

PubMed

Barnes, M P; Ebert, M A

2008-03-01

The concept of electron pencil-beam dose distributions is central to pencil-beam algorithms used in electron beam radiotherapy treatment planning. The Hogstrom algorithm, which is a common algorithm for electron treatment planning, models large electron field dose distributions by the superposition of a series of pencil beam dose distributions. This means that the accurate characterisation of an electron pencil beam is essential for the accuracy of the dose algorithm. The aim of this study was to evaluate a measurement based approach for obtaining electron pencil-beam dose distributions. The primary incentive for the study was the accurate calculation of dose distributions for narrow fields as traditional electron algorithms are generally inaccurate for such geometries. Kodak X-Omat radiographic film was used in a solid water phantom to measure the dose distribution of circular 12 MeV beams from a Varian 21EX linear accelerator. Measurements were made for beams of diameter, 1.5, 2, 4, 8, 16 and 32 mm. A blocked-field technique was used to subtract photon contamination in the beam. The "error function" derived from Fermi-Eyges Multiple Coulomb Scattering (MCS) theory for corresponding square fields was used to fit resulting dose distributions so that extrapolation down to a pencil beam distribution could be made. The Monte Carlo codes, BEAM and EGSnrc were used to simulate the experimental arrangement. The 8 mm beam dose distribution was also measured with TLD-100 microcubes. Agreement between film, TLD and Monte Carlo simulation results were found to be consistent with the spatial resolution used. The study has shown that it is possible to extrapolate narrow electron beam dose distributions down to a pencil beam dose distribution using the error function. However, due to experimental uncertainties and measurement difficulties, Monte Carlo is recommended as the method of choice for characterising electron pencil-beam dose distributions.

Learning control system design based on 2-D theory - An application to parallel link manipulator

NASA Technical Reports Server (NTRS)

Geng, Z.; Carroll, R. L.; Lee, J. D.; Haynes, L. H.

1990-01-01

An approach to iterative learning control system design based on two-dimensional system theory is presented. A two-dimensional model for the iterative learning control system which reveals the connections between learning control systems and two-dimensional system theory is established. A learning control algorithm is proposed, and the convergence of learning using this algorithm is guaranteed by two-dimensional stability. The learning algorithm is applied successfully to the trajectory tracking control problem for a parallel link robot manipulator. The excellent performance of this learning algorithm is demonstrated by the computer simulation results.

Monte Carlo charged-particle tracking and energy deposition on a Lagrangian mesh.

PubMed

Yuan, J; Moses, G A; McKenty, P W

2005-10-01

A Monte Carlo algorithm for alpha particle tracking and energy deposition on a cylindrical computational mesh in a Lagrangian hydrodynamics code used for inertial confinement fusion (ICF) simulations is presented. The straight line approximation is used to follow propagation of "Monte Carlo particles" which represent collections of alpha particles generated from thermonuclear deuterium-tritium (DT) reactions. Energy deposition in the plasma is modeled by the continuous slowing down approximation. The scheme addresses various aspects arising in the coupling of Monte Carlo tracking with Lagrangian hydrodynamics; such as non-orthogonal severely distorted mesh cells, particle relocation on the moving mesh and particle relocation after rezoning. A comparison with the flux-limited multi-group diffusion transport method is presented for a polar direct drive target design for the National Ignition Facility. Simulations show the Monte Carlo transport method predicts about earlier ignition than predicted by the diffusion method, and generates higher hot spot temperature. Nearly linear speed-up is achieved for multi-processor parallel simulations.

QMCPACK : an open source ab initio quantum Monte Carlo package for the electronic structure of atoms, molecules and solids

DOE PAGES

Kim, Jeongnim; Baczewski, Andrew T.; Beaudet, Todd D.; ...

2018-04-19

QMCPACK is an open source quantum Monte Carlo package for ab-initio electronic structure calculations. It supports calculations of metallic and insulating solids, molecules, atoms, and some model Hamiltonians. Implemented real space quantum Monte Carlo algorithms include variational, diffusion, and reptation Monte Carlo. QMCPACK uses Slater-Jastrow type trial wave functions in conjunction with a sophisticated optimizer capable of optimizing tens of thousands of parameters. The orbital space auxiliary field quantum Monte Carlo method is also implemented, enabling cross validation between different highly accurate methods. The code is specifically optimized for calculations with large numbers of electrons on the latest high performancemore » computing architectures, including multicore central processing unit (CPU) and graphical processing unit (GPU) systems. We detail the program’s capabilities, outline its structure, and give examples of its use in current research calculations. The package is available at http://www.qmcpack.org.« less

QMCPACK : an open source ab initio quantum Monte Carlo package for the electronic structure of atoms, molecules and solids

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

Kim, Jeongnim; Baczewski, Andrew T.; Beaudet, Todd D.

QMCPACK is an open source quantum Monte Carlo package for ab-initio electronic structure calculations. It supports calculations of metallic and insulating solids, molecules, atoms, and some model Hamiltonians. Implemented real space quantum Monte Carlo algorithms include variational, diffusion, and reptation Monte Carlo. QMCPACK uses Slater-Jastrow type trial wave functions in conjunction with a sophisticated optimizer capable of optimizing tens of thousands of parameters. The orbital space auxiliary field quantum Monte Carlo method is also implemented, enabling cross validation between different highly accurate methods. The code is specifically optimized for calculations with large numbers of electrons on the latest high performancemore » computing architectures, including multicore central processing unit (CPU) and graphical processing unit (GPU) systems. We detail the program’s capabilities, outline its structure, and give examples of its use in current research calculations. The package is available at http://www.qmcpack.org.« less

Tracking student progress in a game-like physics learning environment with a Monte Carlo Bayesian knowledge tracing model

NASA Astrophysics Data System (ADS)

Gweon, Gey-Hong; Lee, Hee-Sun; Dorsey, Chad; Tinker, Robert; Finzer, William; Damelin, Daniel

2015-03-01

In tracking student learning in on-line learning systems, the Bayesian knowledge tracing (BKT) model is a popular model. However, the model has well-known problems such as the identifiability problem or the empirical degeneracy problem. Understanding of these problems remain unclear and solutions to them remain subjective. Here, we analyze the log data from an online physics learning program with our new model, a Monte Carlo BKT model. With our new approach, we are able to perform a completely unbiased analysis, which can then be used for classifying student learning patterns and performances. Furthermore, a theoretical analysis of the BKT model and our computational work shed new light on the nature of the aforementioned problems. This material is based upon work supported by the National Science Foundation under Grant REC-1147621 and REC-1435470.

Analysis of Monte Carlo accelerated iterative methods for sparse linear systems: Analysis of Monte Carlo accelerated iterative methods for sparse linear systems

DOE PAGES

Benzi, Michele; Evans, Thomas M.; Hamilton, Steven P.; ...

2017-03-05

Here, we consider hybrid deterministic-stochastic iterative algorithms for the solution of large, sparse linear systems. Starting from a convergent splitting of the coefficient matrix, we analyze various types of Monte Carlo acceleration schemes applied to the original preconditioned Richardson (stationary) iteration. We expect that these methods will have considerable potential for resiliency to faults when implemented on massively parallel machines. We also establish sufficient conditions for the convergence of the hybrid schemes, and we investigate different types of preconditioners including sparse approximate inverses. Numerical experiments on linear systems arising from the discretization of partial differential equations are presented.

Active Learning Using Hint Information.

PubMed

Li, Chun-Liang; Ferng, Chun-Sung; Lin, Hsuan-Tien

2015-08-01

The abundance of real-world data and limited labeling budget calls for active learning, an important learning paradigm for reducing human labeling efforts. Many recently developed active learning algorithms consider both uncertainty and representativeness when making querying decisions. However, exploiting representativeness with uncertainty concurrently usually requires tackling sophisticated and challenging learning tasks, such as clustering. In this letter, we propose a new active learning framework, called hinted sampling, which takes both uncertainty and representativeness into account in a simpler way. We design a novel active learning algorithm within the hinted sampling framework with an extended support vector machine. Experimental results validate that the novel active learning algorithm can result in a better and more stable performance than that achieved by state-of-the-art algorithms. We also show that the hinted sampling framework allows improving another active learning algorithm designed from the transductive support vector machine.

A reversible-jump Markov chain Monte Carlo algorithm for 1D inversion of magnetotelluric data

NASA Astrophysics Data System (ADS)

Mandolesi, Eric; Ogaya, Xenia; Campanyà, Joan; Piana Agostinetti, Nicola

2018-04-01

This paper presents a new computer code developed to solve the 1D magnetotelluric (MT) inverse problem using a Bayesian trans-dimensional Markov chain Monte Carlo algorithm. MT data are sensitive to the depth-distribution of rock electric conductivity (or its reciprocal, resistivity). The solution provided is a probability distribution - the so-called posterior probability distribution (PPD) for the conductivity at depth, together with the PPD of the interface depths. The PPD is sampled via a reversible-jump Markov Chain Monte Carlo (rjMcMC) algorithm, using a modified Metropolis-Hastings (MH) rule to accept or discard candidate models along the chains. As the optimal parameterization for the inversion process is generally unknown a trans-dimensional approach is used to allow the dataset itself to indicate the most probable number of parameters needed to sample the PPD. The algorithm is tested against two simulated datasets and a set of MT data acquired in the Clare Basin (County Clare, Ireland). For the simulated datasets the correct number of conductive layers at depth and the associated electrical conductivity values is retrieved, together with reasonable estimates of the uncertainties on the investigated parameters. Results from the inversion of field measurements are compared with results obtained using a deterministic method and with well-log data from a nearby borehole. The PPD is in good agreement with the well-log data, showing as a main structure a high conductive layer associated with the Clare Shale formation. In this study, we demonstrate that our new code go beyond algorithms developend using a linear inversion scheme, as it can be used: (1) to by-pass the subjective choices in the 1D parameterizations, i.e. the number of horizontal layers in the 1D parameterization, and (2) to estimate realistic uncertainties on the retrieved parameters. The algorithm is implemented using a simple MPI approach, where independent chains run on isolated CPU, to take full advantage of parallel computer architectures. In case of a large number of data, a master/slave appoach can be used, where the master CPU samples the parameter space and the slave CPUs compute forward solutions.

A developmental approach to learning causal models for cyber security

NASA Astrophysics Data System (ADS)

Mugan, Jonathan

2013-05-01

To keep pace with our adversaries, we must expand the scope of machine learning and reasoning to address the breadth of possible attacks. One approach is to employ an algorithm to learn a set of causal models that describes the entire cyber network and each host end node. Such a learning algorithm would run continuously on the system and monitor activity in real time. With a set of causal models, the algorithm could anticipate novel attacks, take actions to thwart them, and predict the second-order effects flood of information, and the algorithm would have to determine which streams of that flood were relevant in which situations. This paper will present the results of efforts toward the application of a developmental learning algorithm to the problem of cyber security. The algorithm is modeled on the principles of human developmental learning and is designed to allow an agent to learn about the computer system in which it resides through active exploration. Children are flexible learners who acquire knowledge by actively exploring their environment and making predictions about what they will find,1, 2 and our algorithm is inspired by the work of the developmental psychologist Jean Piaget.3 Piaget described how children construct knowledge in stages and learn new concepts on top of those they already know. Developmental learning allows our algorithm to focus on subsets of the environment that are most helpful for learning given its current knowledge. In experiments, the algorithm was able to learn the conditions for file exfiltration and use that knowledge to protect sensitive files.

Discrete-Time Deterministic $Q$ -Learning: A Novel Convergence Analysis.

PubMed

Wei, Qinglai; Lewis, Frank L; Sun, Qiuye; Yan, Pengfei; Song, Ruizhuo

2017-05-01

In this paper, a novel discrete-time deterministic Q -learning algorithm is developed. In each iteration of the developed Q -learning algorithm, the iterative Q function is updated for all the state and control spaces, instead of updating for a single state and a single control in traditional Q -learning algorithm. A new convergence criterion is established to guarantee that the iterative Q function converges to the optimum, where the convergence criterion of the learning rates for traditional Q -learning algorithms is simplified. During the convergence analysis, the upper and lower bounds of the iterative Q function are analyzed to obtain the convergence criterion, instead of analyzing the iterative Q function itself. For convenience of analysis, the convergence properties for undiscounted case of the deterministic Q -learning algorithm are first developed. Then, considering the discounted factor, the convergence criterion for the discounted case is established. Neural networks are used to approximate the iterative Q function and compute the iterative control law, respectively, for facilitating the implementation of the deterministic Q -learning algorithm. Finally, simulation results and comparisons are given to illustrate the performance of the developed algorithm.

CYBER 200 Applications Seminar

NASA Technical Reports Server (NTRS)

Gary, J. P. (Compiler)

1984-01-01

Applications suited for the CYBER 200 digital computer are discussed. Various areas of application including meteorology, algorithms, fluid dynamics, monte carlo methods, petroleum, electronic circuit simulation, biochemistry, lattice gauge theory, economics and ray tracing are discussed.

Reference hypernetted chain theory for ferrofluid bilayer: Distribution functions compared with Monte Carlo

NASA Astrophysics Data System (ADS)

Polyakov, Evgeny A.; Vorontsov-Velyaminov, Pavel N.

2014-08-01

Properties of ferrofluid bilayer (modeled as a system of two planar layers separated by a distance h and each layer carrying a soft sphere dipolar liquid) are calculated in the framework of inhomogeneous Ornstein-Zernike equations with reference hypernetted chain closure (RHNC). The bridge functions are taken from a soft sphere (1/r12) reference system in the pressure-consistent closure approximation. In order to make the RHNC problem tractable, the angular dependence of the correlation functions is expanded into special orthogonal polynomials according to Lado. The resulting equations are solved using the Newton-GRMES algorithm as implemented in the public-domain solver NITSOL. Orientational densities and pair distribution functions of dipoles are compared with Monte Carlo simulation results. A numerical algorithm for the Fourier-Hankel transform of any positive integer order on a uniform grid is presented.

Exploring Neutrino Oscillation Parameter Space with a Monte Carlo Algorithm

NASA Astrophysics Data System (ADS)

Espejel, Hugo; Ernst, David; Cogswell, Bernadette; Latimer, David

2015-04-01

The χ2 (or likelihood) function for a global analysis of neutrino oscillation data is first calculated as a function of the neutrino mixing parameters. A computational challenge is to obtain the minima or the allowed regions for the mixing parameters. The conventional approach is to calculate the χ2 (or likelihood) function on a grid for a large number of points, and then marginalize over the likelihood function. As the number of parameters increases with the number of neutrinos, making the calculation numerically efficient becomes necessary. We implement a new Monte Carlo algorithm (D. Foreman-Mackey, D. W. Hogg, D. Lang and J. Goodman, Publications of the Astronomical Society of the Pacific, 125 306 (2013)) to determine its computational efficiency at finding the minima and allowed regions. We examine a realistic example to compare the historical and the new methods.

Bayesian forecasting and uncertainty quantifying of stream flows using Metropolis-Hastings Markov Chain Monte Carlo algorithm

NASA Astrophysics Data System (ADS)

Wang, Hongrui; Wang, Cheng; Wang, Ying; Gao, Xiong; Yu, Chen

2017-06-01

This paper presents a Bayesian approach using Metropolis-Hastings Markov Chain Monte Carlo algorithm and applies this method for daily river flow rate forecast and uncertainty quantification for Zhujiachuan River using data collected from Qiaotoubao Gage Station and other 13 gage stations in Zhujiachuan watershed in China. The proposed method is also compared with the conventional maximum likelihood estimation (MLE) for parameter estimation and quantification of associated uncertainties. While the Bayesian method performs similarly in estimating the mean value of daily flow rate, it performs over the conventional MLE method on uncertainty quantification, providing relatively narrower reliable interval than the MLE confidence interval and thus more precise estimation by using the related information from regional gage stations. The Bayesian MCMC method might be more favorable in the uncertainty analysis and risk management.

FitSKIRT: genetic algorithms to automatically fit dusty galaxies with a Monte Carlo radiative transfer code

NASA Astrophysics Data System (ADS)

De Geyter, G.; Baes, M.; Fritz, J.; Camps, P.

2013-02-01

We present FitSKIRT, a method to efficiently fit radiative transfer models to UV/optical images of dusty galaxies. These images have the advantage that they have better spatial resolution compared to FIR/submm data. FitSKIRT uses the GAlib genetic algorithm library to optimize the output of the SKIRT Monte Carlo radiative transfer code. Genetic algorithms prove to be a valuable tool in handling the multi- dimensional search space as well as the noise induced by the random nature of the Monte Carlo radiative transfer code. FitSKIRT is tested on artificial images of a simulated edge-on spiral galaxy, where we gradually increase the number of fitted parameters. We find that we can recover all model parameters, even if all 11 model parameters are left unconstrained. Finally, we apply the FitSKIRT code to a V-band image of the edge-on spiral galaxy NGC 4013. This galaxy has been modeled previously by other authors using different combinations of radiative transfer codes and optimization methods. Given the different models and techniques and the complexity and degeneracies in the parameter space, we find reasonable agreement between the different models. We conclude that the FitSKIRT method allows comparison between different models and geometries in a quantitative manner and minimizes the need of human intervention and biasing. The high level of automation makes it an ideal tool to use on larger sets of observed data.

System Design under Uncertainty: Evolutionary Optimization of the Gravity Probe-B Spacecraft

NASA Technical Reports Server (NTRS)

Pullen, Samuel P.; Parkinson, Bradford W.

1994-01-01

This paper discusses the application of evolutionary random-search algorithms (Simulated Annealing and Genetic Algorithms) to the problem of spacecraft design under performance uncertainty. Traditionally, spacecraft performance uncertainty has been measured by reliability. Published algorithms for reliability optimization are seldom used in practice because they oversimplify reality. The algorithm developed here uses random-search optimization to allow us to model the problem more realistically. Monte Carlo simulations are used to evaluate the objective function for each trial design solution. These methods have been applied to the Gravity Probe-B (GP-B) spacecraft being developed at Stanford University for launch in 1999, Results of the algorithm developed here for GP-13 are shown, and their implications for design optimization by evolutionary algorithms are discussed.

A New Algorithm with Plane Waves and Wavelets for Random Velocity Fields with Many Spatial Scales

NASA Astrophysics Data System (ADS)

Elliott, Frank W.; Majda, Andrew J.

1995-03-01

A new Monte Carlo algorithm for constructing and sampling stationary isotropic Gaussian random fields with power-law energy spectrum, infrared divergence, and fractal self-similar scaling is developed here. The theoretical basis for this algorithm involves the fact that such a random field is well approximated by a superposition of random one-dimensional plane waves involving a fixed finite number of directions. In general each one-dimensional plane wave is the sum of a random shear layer and a random acoustical wave. These one-dimensional random plane waves are then simulated by a wavelet Monte Carlo method for a single space variable developed recently by the authors. The computational results reported in this paper demonstrate remarkable low variance and economical representation of such Gaussian random fields through this new algorithm. In particular, the velocity structure function for an imcorepressible isotropic Gaussian random field in two space dimensions with the Kolmogoroff spectrum can be simulated accurately over 12 decades with only 100 realizations of the algorithm with the scaling exponent accurate to 1.1% and the constant prefactor accurate to 6%; in fact, the exponent of the velocity structure function can be computed over 12 decades within 3.3% with only 10 realizations. Furthermore, only 46,592 active computational elements are utilized in each realization to achieve these results for 12 decades of scaling behavior.

Developing a cosmic ray muon sampling capability for muon tomography and monitoring applications

NASA Astrophysics Data System (ADS)

Chatzidakis, S.; Chrysikopoulou, S.; Tsoukalas, L. H.

2015-12-01

In this study, a cosmic ray muon sampling capability using a phenomenological model that captures the main characteristics of the experimentally measured spectrum coupled with a set of statistical algorithms is developed. The "muon generator" produces muons with zenith angles in the range 0-90° and energies in the range 1-100 GeV and is suitable for Monte Carlo simulations with emphasis on muon tomographic and monitoring applications. The muon energy distribution is described by the Smith and Duller (1959) [35] phenomenological model. Statistical algorithms are then employed for generating random samples. The inverse transform provides a means to generate samples from the muon angular distribution, whereas the Acceptance-Rejection and Metropolis-Hastings algorithms are employed to provide the energy component. The predictions for muon energies 1-60 GeV and zenith angles 0-90° are validated with a series of actual spectrum measurements and with estimates from the software library CRY. The results confirm the validity of the phenomenological model and the applicability of the statistical algorithms to generate polyenergetic-polydirectional muons. The response of the algorithms and the impact of critical parameters on computation time and computed results were investigated. Final output from the proposed "muon generator" is a look-up table that contains the sampled muon angles and energies and can be easily integrated into Monte Carlo particle simulation codes such as Geant4 and MCNP.

A Piagetian Learning Cycle for Introductory Chemical Kinetics.

ERIC Educational Resources Information Center

Batt, Russell H.

1980-01-01

Described is a Piagetian learning cycle based on Monte Carlo modeling of several simple reaction mechanisms. Included are descriptions of learning cycle phases (exploration, invention, and discovery) and four BASIC-PLUS computer programs to be used in the explanation of chemical reacting systems. (Author/DS)

A study on the performance comparison of metaheuristic algorithms on the learning of neural networks

NASA Astrophysics Data System (ADS)

Lai, Kee Huong; Zainuddin, Zarita; Ong, Pauline

2017-08-01

The learning or training process of neural networks entails the task of finding the most optimal set of parameters, which includes translation vectors, dilation parameter, synaptic weights, and bias terms. Apart from the traditional gradient descent-based methods, metaheuristic methods can also be used for this learning purpose. Since the inception of genetic algorithm half a century ago, the last decade witnessed the explosion of a variety of novel metaheuristic algorithms, such as harmony search algorithm, bat algorithm, and whale optimization algorithm. Despite the proof of the no free lunch theorem in the discipline of optimization, a survey in the literature of machine learning gives contrasting results. Some researchers report that certain metaheuristic algorithms are superior to the others, whereas some others argue that different metaheuristic algorithms give comparable performance. As such, this paper aims to investigate if a certain metaheuristic algorithm will outperform the other algorithms. In this work, three metaheuristic algorithms, namely genetic algorithms, particle swarm optimization, and harmony search algorithm are considered. The algorithms are incorporated in the learning of neural networks and their classification results on the benchmark UCI machine learning data sets are compared. It is found that all three metaheuristic algorithms give similar and comparable performance, as captured in the average overall classification accuracy. The results corroborate the findings reported in the works done by previous researchers. Several recommendations are given, which include the need of statistical analysis to verify the results and further theoretical works to support the obtained empirical results.

Fast GPU-based Monte Carlo simulations for LDR prostate brachytherapy.

PubMed

Bonenfant, Éric; Magnoux, Vincent; Hissoiny, Sami; Ozell, Benoît; Beaulieu, Luc; Després, Philippe

2015-07-07

The aim of this study was to evaluate the potential of bGPUMCD, a Monte Carlo algorithm executed on Graphics Processing Units (GPUs), for fast dose calculations in permanent prostate implant dosimetry. It also aimed to validate a low dose rate brachytherapy source in terms of TG-43 metrics and to use this source to compute dose distributions for permanent prostate implant in very short times. The physics of bGPUMCD was reviewed and extended to include Rayleigh scattering and fluorescence from photoelectric interactions for all materials involved. The radial and anisotropy functions were obtained for the Nucletron SelectSeed in TG-43 conditions. These functions were compared to those found in the MD Anderson Imaging and Radiation Oncology Core brachytherapy source registry which are considered the TG-43 reference values. After appropriate calibration of the source, permanent prostate implant dose distributions were calculated for four patients and compared to an already validated Geant4 algorithm. The radial function calculated from bGPUMCD showed excellent agreement (differences within 1.3%) with TG-43 accepted values. The anisotropy functions at r = 1 cm and r = 4 cm were within 2% of TG-43 values for angles over 17.5°. For permanent prostate implants, Monte Carlo-based dose distributions with a statistical uncertainty of 1% or less for the target volume were obtained in 30 s or less for 1 × 1 × 1 mm(3) calculation grids. Dosimetric indices were very similar (within 2.7%) to those obtained with a validated, independent Monte Carlo code (Geant4) performing the calculations for the same cases in a much longer time (tens of minutes to more than a hour). bGPUMCD is a promising code that lets envision the use of Monte Carlo techniques in a clinical environment, with sub-minute execution times on a standard workstation. Future work will explore the use of this code with an inverse planning method to provide a complete Monte Carlo-based planning solution.

Fast GPU-based Monte Carlo simulations for LDR prostate brachytherapy

NASA Astrophysics Data System (ADS)

Bonenfant, Éric; Magnoux, Vincent; Hissoiny, Sami; Ozell, Benoît; Beaulieu, Luc; Després, Philippe

2015-07-01

The aim of this study was to evaluate the potential of bGPUMCD, a Monte Carlo algorithm executed on Graphics Processing Units (GPUs), for fast dose calculations in permanent prostate implant dosimetry. It also aimed to validate a low dose rate brachytherapy source in terms of TG-43 metrics and to use this source to compute dose distributions for permanent prostate implant in very short times. The physics of bGPUMCD was reviewed and extended to include Rayleigh scattering and fluorescence from photoelectric interactions for all materials involved. The radial and anisotropy functions were obtained for the Nucletron SelectSeed in TG-43 conditions. These functions were compared to those found in the MD Anderson Imaging and Radiation Oncology Core brachytherapy source registry which are considered the TG-43 reference values. After appropriate calibration of the source, permanent prostate implant dose distributions were calculated for four patients and compared to an already validated Geant4 algorithm. The radial function calculated from bGPUMCD showed excellent agreement (differences within 1.3%) with TG-43 accepted values. The anisotropy functions at r = 1 cm and r = 4 cm were within 2% of TG-43 values for angles over 17.5°. For permanent prostate implants, Monte Carlo-based dose distributions with a statistical uncertainty of 1% or less for the target volume were obtained in 30 s or less for 1 × 1 × 1 mm3 calculation grids. Dosimetric indices were very similar (within 2.7%) to those obtained with a validated, independent Monte Carlo code (Geant4) performing the calculations for the same cases in a much longer time (tens of minutes to more than a hour). bGPUMCD is a promising code that lets envision the use of Monte Carlo techniques in a clinical environment, with sub-minute execution times on a standard workstation. Future work will explore the use of this code with an inverse planning method to provide a complete Monte Carlo-based planning solution.

SU-E-T-171: Evaluation of the Analytical Anisotropic Algorithm in a Small Finger Joint Phantom Using Monte Carlo Simulation

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

Chow, J; Owrangi, A; Jiang, R

2014-06-01

Purpose: This study investigated the performance of the anisotropic analytical algorithm (AAA) in dose calculation in radiotherapy concerning a small finger joint. Monte Carlo simulation (EGSnrc code) was used in this dosimetric evaluation. Methods: Heterogeneous finger joint phantom containing a vertical water layer (bone joint or cartilage) sandwiched by two bones with dimension 2 × 2 × 2 cm{sup 3} was irradiated by the 6 MV photon beams (field size = 4 × 4 cm{sup 2}). The central beam axis was along the length of the bone joint and the isocenter was set to the center of the joint. Themore » joint width and beam angle were varied from 0.5–2 mm and 0°–15°, respectively. Depth doses were calculated using the AAA and DOSXYZnrc. For dosimetric comparison and normalization, dose calculations were repeated in water phantom using the same beam geometry. Results: Our AAA and Monte Carlo results showed that the AAA underestimated the joint doses by 10%–20%, and could not predict joint dose variation with changes of joint width and beam angle. The calculated bone dose enhancement for the AAA was lower than Monte Carlo and the depth of maximum dose for the phantom was smaller than that for the water phantom. From Monte Carlo results, there was a decrease of joint dose as its width increased. This reflected the smaller the joint width, the more the bone scatter contributed to the depth dose. Moreover, the joint dose was found slightly decreased with an increase of beam angle. Conclusion: The AAA could not handle variations of joint dose well with changes of joint width and beam angle based on our finger joint phantom. Monte Carlo results showed that the joint dose decreased with increase of joint width and beam angle. This dosimetry comparison should be useful to radiation staff in radiotherapy related to small bone joint.« less

Monte Carlo-based Reconstruction in Water Cherenkov Detectors using Chroma

NASA Astrophysics Data System (ADS)

Seibert, Stanley; Latorre, Anthony

2012-03-01

We demonstrate the feasibility of event reconstruction---including position, direction, energy and particle identification---in water Cherenkov detectors with a purely Monte Carlo-based method. Using a fast optical Monte Carlo package we have written, called Chroma, in combination with several variance reduction techniques, we can estimate the value of a likelihood function for an arbitrary event hypothesis. The likelihood can then be maximized over the parameter space of interest using a form of gradient descent designed for stochastic functions. Although slower than more traditional reconstruction algorithms, this completely Monte Carlo-based technique is universal and can be applied to a detector of any size or shape, which is a major advantage during the design phase of an experiment. As a specific example, we focus on reconstruction results from a simulation of the 200 kiloton water Cherenkov far detector option for LBNE.

North Alabama Lightning Mapping Array (LMA): VHF Source Retrieval Algorithm and Error Analyses

NASA Technical Reports Server (NTRS)

Koshak, W. J.; Solakiewicz, R. J.; Blakeslee, R. J.; Goodman, S. J.; Christian, H. J.; Hall, J.; Bailey, J.; Krider, E. P.; Bateman, M. G.; Boccippio, D.

2003-01-01

Two approaches are used to characterize how accurately the North Alabama Lightning Mapping Array (LMA) is able to locate lightning VHF sources in space and in time. The first method uses a Monte Carlo computer simulation to estimate source retrieval errors. The simulation applies a VHF source retrieval algorithm that was recently developed at the NASA Marshall Space Flight Center (MSFC) and that is similar, but not identical to, the standard New Mexico Tech retrieval algorithm. The second method uses a purely theoretical technique (i.e., chi-squared Curvature Matrix Theory) to estimate retrieval errors. Both methods assume that the LMA system has an overall rms timing error of 50 ns, but all other possible errors (e.g., multiple sources per retrieval attempt) are neglected. The detailed spatial distributions of retrieval errors are provided. Given that the two methods are completely independent of one another, it is shown that they provide remarkably similar results. However, for many source locations, the Curvature Matrix Theory produces larger altitude error estimates than the (more realistic) Monte Carlo simulation.

Commissioning and initial acceptance tests for a commercial convolution dose calculation algorithm for radiotherapy treatment planning in comparison with Monte Carlo simulation and measurement

PubMed Central

Moradi, Farhad; Mahdavi, Seyed Rabi; Mostaar, Ahmad; Motamedi, Mohsen

2012-01-01

In this study the commissioning of a dose calculation algorithm in a currently used treatment planning system was performed and the calculation accuracy of two available methods in the treatment planning system i.e., collapsed cone convolution (CCC) and equivalent tissue air ratio (ETAR) was verified in tissue heterogeneities. For this purpose an inhomogeneous phantom (IMRT thorax phantom) was used and dose curves obtained by the TPS (treatment planning system) were compared with experimental measurements and Monte Carlo (MCNP code) simulation. Dose measurements were performed by using EDR2 radiographic films within the phantom. Dose difference (DD) between experimental results and two calculation methods was obtained. Results indicate maximum difference of 12% in the lung and 3% in the bone tissue of the phantom between two methods and the CCC algorithm shows more accurate depth dose curves in tissue heterogeneities. Simulation results show the accurate dose estimation by MCNP4C in soft tissue region of the phantom and also better results than ETAR method in bone and lung tissues. PMID:22973081

A kMC-MD method with generalized move-sets for the simulation of folding of α-helical and β-stranded peptides.

PubMed

Peter, Emanuel K; Pivkin, Igor V; Shea, Joan-Emma

2015-04-14

In Monte-Carlo simulations of protein folding, pathways and folding times depend on the appropriate choice of the Monte-Carlo move or process path. We developed a generalized set of process paths for a hybrid kinetic Monte Carlo-Molecular dynamics algorithm, which makes use of a novel constant time-update and allows formation of α-helical and β-stranded secondary structures. We apply our new algorithm to the folding of 3 different proteins: TrpCage, GB1, and TrpZip4. All three systems are seen to fold within the range of the experimental folding times. For the β-hairpins, we observe that loop formation is the rate-determining process followed by collapse and formation of the native core. Cluster analysis of both peptides reveals that GB1 folds with equal likelihood along a zipper or a hydrophobic collapse mechanism, while TrpZip4 follows primarily a zipper pathway. The difference observed in the folding behavior of the two proteins can be attributed to the different arrangements of their hydrophobic core, strongly packed, and dry in case of TrpZip4, and partially hydrated in the case of GB1.

A junction-tree based learning algorithm to optimize network wide traffic control: A coordinated multi-agent framework

DOE PAGES

Zhu, Feng; Aziz, H. M. Abdul; Qian, Xinwu; ...

2015-01-31

Our study develops a novel reinforcement learning algorithm for the challenging coordinated signal control problem. Traffic signals are modeled as intelligent agents interacting with the stochastic traffic environment. The model is built on the framework of coordinated reinforcement learning. The Junction Tree Algorithm (JTA) based reinforcement learning is proposed to obtain an exact inference of the best joint actions for all the coordinated intersections. Moreover, the algorithm is implemented and tested with a network containing 18 signalized intersections in VISSIM. Finally, our results show that the JTA based algorithm outperforms independent learning (Q-learning), real-time adaptive learning, and fixed timing plansmore » in terms of average delay, number of stops, and vehicular emissions at the network level.« less

First-Order Phase Transition in the Quantum Adiabatic Algorithm

DTIC Science & Technology

2010-01-14

London) 400, 133 (1999). [19] T. Jörg, F. Krzakala, G . Semerjian, and F. Zamponi, arXiv:0911.3438. PRL 104, 020502 (2010) P HY S I CA L R EV I EW LE T T E R S week ending 15 JANUARY 2010 020502-4 ...Box 12211 Research Triangle Park, NC 27709-2211 15. SUBJECT TERMS Quantum Adiabatic Algorithm, Monte Carlo, Quantum Phase Transition A. P . Young, V...documentation. Approved for public release; distribution is unlimited. ... 56290.2-PH-QC First-Order Phase Transition in the Quantum Adiabatic Algorithm A. P

Vega roll and attitude control system algorithms trade-off study

NASA Astrophysics Data System (ADS)

Paulino, N.; Cuciniello, G.; Cruciani, I.; Corraro, F.; Spallotta, D.; Nebula, F.

2013-12-01

This paper describes the trade-off study for the selection of the most suitable algorithms for the Roll and Attitude Control System (RACS) within the FPS-A program, aimed at developing the new Flight Program Software of VEGA Launcher. Two algorithms were analyzed: Switching Lines (SL) and Quaternion Feedback Regulation. Using a development simulation tool that models two critical flight phases (Long Coasting Phase (LCP) and Payload Release (PLR) Phase), both algorithms were assessed with Monte Carlo batch simulations for both of the phases. The statistical outcomes of the results demonstrate a 100 percent success rate for Quaternion Feedback Regulation, and support the choice of this method.

Asteroid mass estimation using Markov-Chain Monte Carlo techniques

NASA Astrophysics Data System (ADS)

Siltala, Lauri; Granvik, Mikael

2016-10-01

Estimates for asteroid masses are based on their gravitational perturbations on the orbits of other objects such as Mars, spacecraft, or other asteroids and/or their satellites. In the case of asteroid-asteroid perturbations, this leads to a 13-dimensional inverse problem where the aim is to derive the mass of the perturbing asteroid and six orbital elements for both the perturbing asteroid and the test asteroid using astrometric observations. We have developed and implemented three different mass estimation algorithms utilizing asteroid-asteroid perturbations into the OpenOrb asteroid-orbit-computation software: the very rough 'marching' approximation, in which the asteroid orbits are fixed at a given epoch, reducing the problem to a one-dimensional estimation of the mass, an implementation of the Nelder-Mead simplex method, and most significantly, a Markov-Chain Monte Carlo (MCMC) approach. We will introduce each of these algorithms with particular focus on the MCMC algorithm, and present example results for both synthetic and real data. Our results agree with the published mass estimates, but suggest that the published uncertainties may be misleading as a consequence of using linearized mass-estimation methods. Finally, we discuss remaining challenges with the algorithms as well as future plans, particularly in connection with ESA's Gaia mission.

Assessment of mean-field microkinetic models for CO methanation on stepped metal surfaces using accelerated kinetic Monte Carlo

NASA Astrophysics Data System (ADS)

Andersen, Mie; Plaisance, Craig P.; Reuter, Karsten

2017-10-01

First-principles screening studies aimed at predicting the catalytic activity of transition metal (TM) catalysts have traditionally been based on mean-field (MF) microkinetic models, which neglect the effect of spatial correlations in the adsorbate layer. Here we critically assess the accuracy of such models for the specific case of CO methanation over stepped metals by comparing to spatially resolved kinetic Monte Carlo (kMC) simulations. We find that the typical low diffusion barriers offered by metal surfaces can be significantly increased at step sites, which results in persisting correlations in the adsorbate layer. As a consequence, MF models may overestimate the catalytic activity of TM catalysts by several orders of magnitude. The potential higher accuracy of kMC models comes at a higher computational cost, which can be especially challenging for surface reactions on metals due to a large disparity in the time scales of different processes. In order to overcome this issue, we implement and test a recently developed algorithm for achieving temporal acceleration of kMC simulations. While the algorithm overall performs quite well, we identify some challenging cases which may lead to a breakdown of acceleration algorithms and discuss possible directions for future algorithm development.

Validation of a track repeating algorithm for intensity modulated proton therapy: clinical cases study

NASA Astrophysics Data System (ADS)

Yepes, Pablo P.; Eley, John G.; Liu, Amy; Mirkovic, Dragan; Randeniya, Sharmalee; Titt, Uwe; Mohan, Radhe

2016-04-01

Monte Carlo (MC) methods are acknowledged as the most accurate technique to calculate dose distributions. However, due its lengthy calculation times, they are difficult to utilize in the clinic or for large retrospective studies. Track-repeating algorithms, based on MC-generated particle track data in water, accelerate dose calculations substantially, while essentially preserving the accuracy of MC. In this study, we present the validation of an efficient dose calculation algorithm for intensity modulated proton therapy, the fast dose calculator (FDC), based on a track-repeating technique. We validated the FDC algorithm for 23 patients, which included 7 brain, 6 head-and-neck, 5 lung, 1 spine, 1 pelvis and 3 prostate cases. For validation, we compared FDC-generated dose distributions with those from a full-fledged Monte Carlo based on GEANT4 (G4). We compared dose-volume-histograms, 3D-gamma-indices and analyzed a series of dosimetric indices. More than 99% of the voxels in the voxelized phantoms describing the patients have a gamma-index smaller than unity for the 2%/2 mm criteria. In addition the difference relative to the prescribed dose between the dosimetric indices calculated with FDC and G4 is less than 1%. FDC reduces the calculation times from 5 ms per proton to around 5 μs.

Monte Carlo MP2 on Many Graphical Processing Units.

PubMed

Doran, Alexander E; Hirata, So

2016-10-11

In the Monte Carlo second-order many-body perturbation (MC-MP2) method, the long sum-of-product matrix expression of the MP2 energy, whose literal evaluation may be poorly scalable, is recast into a single high-dimensional integral of functions of electron pair coordinates, which is evaluated by the scalable method of Monte Carlo integration. The sampling efficiency is further accelerated by the redundant-walker algorithm, which allows a maximal reuse of electron pairs. Here, a multitude of graphical processing units (GPUs) offers a uniquely ideal platform to expose multilevel parallelism: fine-grain data-parallelism for the redundant-walker algorithm in which millions of threads compute and share orbital amplitudes on each GPU; coarse-grain instruction-parallelism for near-independent Monte Carlo integrations on many GPUs with few and infrequent interprocessor communications. While the efficiency boost by the redundant-walker algorithm on central processing units (CPUs) grows linearly with the number of electron pairs and tends to saturate when the latter exceeds the number of orbitals, on a GPU it grows quadratically before it increases linearly and then eventually saturates at a much larger number of pairs. This is because the orbital constructions are nearly perfectly parallelized on a GPU and thus completed in a near-constant time regardless of the number of pairs. In consequence, an MC-MP2/cc-pVDZ calculation of a benzene dimer is 2700 times faster on 256 GPUs (using 2048 electron pairs) than on two CPUs, each with 8 cores (which can use only up to 256 pairs effectively). We also numerically determine that the cost to achieve a given relative statistical uncertainty in an MC-MP2 energy increases as O(n 3 ) or better with system size n, which may be compared with the O(n 5 ) scaling of the conventional implementation of deterministic MP2. We thus establish the scalability of MC-MP2 with both system and computer sizes.

An Improved Method of Heterogeneity Compensation for the Convolution / Superposition Algorithm

NASA Astrophysics Data System (ADS)

Jacques, Robert; McNutt, Todd

2014-03-01

Purpose: To improve the accuracy of convolution/superposition (C/S) in heterogeneous material by developing a new algorithm: heterogeneity compensated superposition (HCS). Methods: C/S has proven to be a good estimator of the dose deposited in a homogeneous volume. However, near heterogeneities electron disequilibrium occurs, leading to the faster fall-off and re-buildup of dose. We propose to filter the actual patient density in a position and direction sensitive manner, allowing the dose deposited near interfaces to be increased or decreased relative to C/S. We implemented the effective density function as a multivariate first-order recursive filter and incorporated it into GPU-accelerated, multi-energetic C/S implementation. We compared HCS against C/S using the ICCR 2000 Monte-Carlo accuracy benchmark, 23 similar accuracy benchmarks and 5 patient cases. Results: Multi-energetic HCS increased the dosimetric accuracy for the vast majority of voxels; in many cases near Monte-Carlo results were achieved. We defined the per-voxel error, %|mm, as the minimum of the distance to agreement in mm and the dosimetric percentage error relative to the maximum MC dose. HCS improved the average mean error by 0.79 %|mm for the patient volumes; reducing the average mean error from 1.93 %|mm to 1.14 %|mm. Very low densities (i.e. < 0.1 g / cm3) remained problematic, but may be solvable with a better filter function. Conclusions: HCS improved upon C/S's density scaled heterogeneity correction with a position and direction sensitive density filter. This method significantly improved the accuracy of the GPU based algorithm reaching the accuracy levels of Monte Carlo based methods with performance in a few tenths of seconds per beam. Acknowledgement: Funding for this research was provided by the NSF Cooperative Agreement EEC9731748, Elekta / IMPAC Medical Systems, Inc. and the Johns Hopkins University. James Satterthwaite provided the Monte Carlo benchmark simulations.

TU-F-CAMPUS-T-05: A Cloud-Based Monte Carlo Dose Calculation for Electron Cutout Factors

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

Mitchell, T; Bush, K

Purpose: For electron cutouts of smaller sizes, it is necessary to verify electron cutout factors due to perturbations in electron scattering. Often, this requires a physical measurement using a small ion chamber, diode, or film. The purpose of this study is to develop a fast Monte Carlo based dose calculation framework that requires only a smart phone photograph of the cutout and specification of the SSD and energy to determine the electron cutout factor, with the ultimate goal of making this cloud-based calculation widely available to the medical physics community. Methods: The algorithm uses a pattern recognition technique to identifymore » the corners of the cutout in the photograph as shown in Figure 1. It then corrects for variations in perspective, scaling, and translation of the photograph introduced by the user’s positioning of the camera. Blob detection is used to identify the portions of the cutout which comprise the aperture and the portions which are cutout material. This information is then used define physical densities of the voxels used in the Monte Carlo dose calculation algorithm as shown in Figure 2, and select a particle source from a pre-computed library of phase-spaces scored above the cutout. The electron cutout factor is obtained by taking a ratio of the maximum dose delivered with the cutout in place to the dose delivered under calibration/reference conditions. Results: The algorithm has been shown to successfully identify all necessary features of the electron cutout to perform the calculation. Subsequent testing will be performed to compare the Monte Carlo results with a physical measurement. Conclusion: A simple, cloud-based method of calculating electron cutout factors could eliminate the need for physical measurements and substantially reduce the time required to properly assure accurate dose delivery.« less

An analysis dictionary learning algorithm under a noisy data model with orthogonality constraint.

PubMed

Zhang, Ye; Yu, Tenglong; Wang, Wenwu

2014-01-01

Two common problems are often encountered in analysis dictionary learning (ADL) algorithms. The first one is that the original clean signals for learning the dictionary are assumed to be known, which otherwise need to be estimated from noisy measurements. This, however, renders a computationally slow optimization process and potentially unreliable estimation (if the noise level is high), as represented by the Analysis K-SVD (AK-SVD) algorithm. The other problem is the trivial solution to the dictionary, for example, the null dictionary matrix that may be given by a dictionary learning algorithm, as discussed in the learning overcomplete sparsifying transform (LOST) algorithm. Here we propose a novel optimization model and an iterative algorithm to learn the analysis dictionary, where we directly employ the observed data to compute the approximate analysis sparse representation of the original signals (leading to a fast optimization procedure) and enforce an orthogonality constraint on the optimization criterion to avoid the trivial solutions. Experiments demonstrate the competitive performance of the proposed algorithm as compared with three baselines, namely, the AK-SVD, LOST, and NAAOLA algorithms.

Remediating radium contaminated legacy sites: Advances made through machine learning in routine monitoring of "hot" particles.

PubMed

Varley, Adam; Tyler, Andrew; Smith, Leslie; Dale, Paul; Davies, Mike

2015-07-15

The extensive use of radium during the 20th century for industrial, military and pharmaceutical purposes has led to a large number of contaminated legacy sites across Europe and North America. Sites that pose a high risk to the general public can present expensive and long-term remediation projects. Often the most pragmatic remediation approach is through routine monitoring operating gamma-ray detectors to identify, in real-time, the signal from the most hazardous heterogeneous contamination (hot particles); thus facilitating their removal and safe disposal. However, current detection systems do not fully utilise all spectral information resulting in low detection rates and ultimately an increased risk to the human health. The aim of this study was to establish an optimised detector-algorithm combination. To achieve this, field data was collected using two handheld detectors (sodium iodide and lanthanum bromide) and a number of Monte Carlo simulated hot particles were randomly injected into the field data. This allowed for the detection rate of conventional deterministic (gross counts) and machine learning (neural networks and support vector machines) algorithms to be assessed. The results demonstrated that a Neural Network operated on a sodium iodide detector provided the best detection capability. Compared to deterministic approaches, this optimised detection system could detect a hot particle on average 10cm deeper into the soil column or with half of the activity at the same depth. It was also found that noise presented by internal contamination restricted lanthanum bromide for this application. Copyright © 2015. Published by Elsevier B.V.

The Monte Carlo code MCPTV--Monte Carlo dose calculation in radiation therapy with carbon ions.

PubMed

Karg, Juergen; Speer, Stefan; Schmidt, Manfred; Mueller, Reinhold

2010-07-07

The Monte Carlo code MCPTV is presented. MCPTV is designed for dose calculation in treatment planning in radiation therapy with particles and especially carbon ions. MCPTV has a voxel-based concept and can perform a fast calculation of the dose distribution on patient CT data. Material and density information from CT are taken into account. Electromagnetic and nuclear interactions are implemented. Furthermore the algorithm gives information about the particle spectra and the energy deposition in each voxel. This can be used to calculate the relative biological effectiveness (RBE) for each voxel. Depth dose distributions are compared to experimental data giving good agreement. A clinical example is shown to demonstrate the capabilities of the MCPTV dose calculation.

Split Orthogonal Group: A Guiding Principle for Sign-Problem-Free Fermionic Simulations

NASA Astrophysics Data System (ADS)

Wang, Lei; Liu, Ye-Hua; Iazzi, Mauro; Troyer, Matthias; Harcos, Gergely

2015-12-01

We present a guiding principle for designing fermionic Hamiltonians and quantum Monte Carlo (QMC) methods that are free from the infamous sign problem by exploiting the Lie groups and Lie algebras that appear naturally in the Monte Carlo weight of fermionic QMC simulations. Specifically, rigorous mathematical constraints on the determinants involving matrices that lie in the split orthogonal group provide a guideline for sign-free simulations of fermionic models on bipartite lattices. This guiding principle not only unifies the recent solutions of the sign problem based on the continuous-time quantum Monte Carlo methods and the Majorana representation, but also suggests new efficient algorithms to simulate physical systems that were previously prohibitive because of the sign problem.

Active learning for clinical text classification: is it better than random sampling?

PubMed

Figueroa, Rosa L; Zeng-Treitler, Qing; Ngo, Long H; Goryachev, Sergey; Wiechmann, Eduardo P

2012-01-01

This study explores active learning algorithms as a way to reduce the requirements for large training sets in medical text classification tasks. Three existing active learning algorithms (distance-based (DIST), diversity-based (DIV), and a combination of both (CMB)) were used to classify text from five datasets. The performance of these algorithms was compared to that of passive learning on the five datasets. We then conducted a novel investigation of the interaction between dataset characteristics and the performance results. Classification accuracy and area under receiver operating characteristics (ROC) curves for each algorithm at different sample sizes were generated. The performance of active learning algorithms was compared with that of passive learning using a weighted mean of paired differences. To determine why the performance varies on different datasets, we measured the diversity and uncertainty of each dataset using relative entropy and correlated the results with the performance differences. The DIST and CMB algorithms performed better than passive learning. With a statistical significance level set at 0.05, DIST outperformed passive learning in all five datasets, while CMB was found to be better than passive learning in four datasets. We found strong correlations between the dataset diversity and the DIV performance, as well as the dataset uncertainty and the performance of the DIST algorithm. For medical text classification, appropriate active learning algorithms can yield performance comparable to that of passive learning with considerably smaller training sets. In particular, our results suggest that DIV performs better on data with higher diversity and DIST on data with lower uncertainty.

Active learning for clinical text classification: is it better than random sampling?

PubMed Central

Figueroa, Rosa L; Ngo, Long H; Goryachev, Sergey; Wiechmann, Eduardo P

2012-01-01

Objective This study explores active learning algorithms as a way to reduce the requirements for large training sets in medical text classification tasks. Design Three existing active learning algorithms (distance-based (DIST), diversity-based (DIV), and a combination of both (CMB)) were used to classify text from five datasets. The performance of these algorithms was compared to that of passive learning on the five datasets. We then conducted a novel investigation of the interaction between dataset characteristics and the performance results. Measurements Classification accuracy and area under receiver operating characteristics (ROC) curves for each algorithm at different sample sizes were generated. The performance of active learning algorithms was compared with that of passive learning using a weighted mean of paired differences. To determine why the performance varies on different datasets, we measured the diversity and uncertainty of each dataset using relative entropy and correlated the results with the performance differences. Results The DIST and CMB algorithms performed better than passive learning. With a statistical significance level set at 0.05, DIST outperformed passive learning in all five datasets, while CMB was found to be better than passive learning in four datasets. We found strong correlations between the dataset diversity and the DIV performance, as well as the dataset uncertainty and the performance of the DIST algorithm. Conclusion For medical text classification, appropriate active learning algorithms can yield performance comparable to that of passive learning with considerably smaller training sets. In particular, our results suggest that DIV performs better on data with higher diversity and DIST on data with lower uncertainty. PMID:22707743

Machine Learning Methods for Attack Detection in the Smart Grid.

PubMed

Ozay, Mete; Esnaola, Inaki; Yarman Vural, Fatos Tunay; Kulkarni, Sanjeev R; Poor, H Vincent

2016-08-01

Attack detection problems in the smart grid are posed as statistical learning problems for different attack scenarios in which the measurements are observed in batch or online settings. In this approach, machine learning algorithms are used to classify measurements as being either secure or attacked. An attack detection framework is provided to exploit any available prior knowledge about the system and surmount constraints arising from the sparse structure of the problem in the proposed approach. Well-known batch and online learning algorithms (supervised and semisupervised) are employed with decision- and feature-level fusion to model the attack detection problem. The relationships between statistical and geometric properties of attack vectors employed in the attack scenarios and learning algorithms are analyzed to detect unobservable attacks using statistical learning methods. The proposed algorithms are examined on various IEEE test systems. Experimental analyses show that machine learning algorithms can detect attacks with performances higher than attack detection algorithms that employ state vector estimation methods in the proposed attack detection framework.

A Locality-Constrained and Label Embedding Dictionary Learning Algorithm for Image Classification.

PubMed

Zhengming Li; Zhihui Lai; Yong Xu; Jian Yang; Zhang, David

2017-02-01

Locality and label information of training samples play an important role in image classification. However, previous dictionary learning algorithms do not take the locality and label information of atoms into account together in the learning process, and thus their performance is limited. In this paper, a discriminative dictionary learning algorithm, called the locality-constrained and label embedding dictionary learning (LCLE-DL) algorithm, was proposed for image classification. First, the locality information was preserved using the graph Laplacian matrix of the learned dictionary instead of the conventional one derived from the training samples. Then, the label embedding term was constructed using the label information of atoms instead of the classification error term, which contained discriminating information of the learned dictionary. The optimal coding coefficients derived by the locality-based and label-based reconstruction were effective for image classification. Experimental results demonstrated that the LCLE-DL algorithm can achieve better performance than some state-of-the-art algorithms.

Fidelity-Based Ant Colony Algorithm with Q-learning of Quantum System

NASA Astrophysics Data System (ADS)

Liao, Qin; Guo, Ying; Tu, Yifeng; Zhang, Hang

2018-03-01

Quantum ant colony algorithm (ACA) has potential applications in quantum information processing, such as solutions of traveling salesman problem, zero-one knapsack problem, robot route planning problem, and so on. To shorten the search time of the ACA, we suggest the fidelity-based ant colony algorithm (FACA) for the control of quantum system. Motivated by structure of the Q-learning algorithm, we demonstrate the combination of a FACA with the Q-learning algorithm and suggest the design of a fidelity-based ant colony algorithm with the Q-learning to improve the performance of the FACA in a spin-1/2 quantum system. The numeric simulation results show that the FACA with the Q-learning can efficiently avoid trapping into local optimal policies and increase the speed of convergence process of quantum system.

Fidelity-Based Ant Colony Algorithm with Q-learning of Quantum System

NASA Astrophysics Data System (ADS)

Liao, Qin; Guo, Ying; Tu, Yifeng; Zhang, Hang

2017-12-01

Quantum ant colony algorithm (ACA) has potential applications in quantum information processing, such as solutions of traveling salesman problem, zero-one knapsack problem, robot route planning problem, and so on. To shorten the search time of the ACA, we suggest the fidelity-based ant colony algorithm (FACA) for the control of quantum system. Motivated by structure of the Q-learning algorithm, we demonstrate the combination of a FACA with the Q-learning algorithm and suggest the design of a fidelity-based ant colony algorithm with the Q-learning to improve the performance of the FACA in a spin-1/2 quantum system. The numeric simulation results show that the FACA with the Q-learning can efficiently avoid trapping into local optimal policies and increase the speed of convergence process of quantum system.

Human resource recommendation algorithm based on ensemble learning and Spark

NASA Astrophysics Data System (ADS)

Cong, Zihan; Zhang, Xingming; Wang, Haoxiang; Xu, Hongjie

2017-08-01

Aiming at the problem of “information overload” in the human resources industry, this paper proposes a human resource recommendation algorithm based on Ensemble Learning. The algorithm considers the characteristics and behaviours of both job seeker and job features in the real business circumstance. Firstly, the algorithm uses two ensemble learning methods-Bagging and Boosting. The outputs from both learning methods are then merged to form user interest model. Based on user interest model, job recommendation can be extracted for users. The algorithm is implemented as a parallelized recommendation system on Spark. A set of experiments have been done and analysed. The proposed algorithm achieves significant improvement in accuracy, recall rate and coverage, compared with recommendation algorithms such as UserCF and ItemCF.

A Coulomb collision algorithm for weighted particle simulations

NASA Technical Reports Server (NTRS)

Miller, Ronald H.; Combi, Michael R.

1994-01-01

A binary Coulomb collision algorithm is developed for weighted particle simulations employing Monte Carlo techniques. Charged particles within a given spatial grid cell are pair-wise scattered, explicitly conserving momentum and implicitly conserving energy. A similar algorithm developed by Takizuka and Abe (1977) conserves momentum and energy provided the particles are unweighted (each particle representing equal fractions of the total particle density). If applied as is to simulations incorporating weighted particles, the plasma temperatures equilibrate to an incorrect temperature, as compared to theory. Using the appropriate pairing statistics, a Coulomb collision algorithm is developed for weighted particles. The algorithm conserves energy and momentum and produces the appropriate relaxation time scales as compared to theoretical predictions. Such an algorithm is necessary for future work studying self-consistent multi-species kinetic transport.

Approximate Bayesian Computation by Subset Simulation using hierarchical state-space models

NASA Astrophysics Data System (ADS)

Vakilzadeh, Majid K.; Huang, Yong; Beck, James L.; Abrahamsson, Thomas

2017-02-01

A new multi-level Markov Chain Monte Carlo algorithm for Approximate Bayesian Computation, ABC-SubSim, has recently appeared that exploits the Subset Simulation method for efficient rare-event simulation. ABC-SubSim adaptively creates a nested decreasing sequence of data-approximating regions in the output space that correspond to increasingly closer approximations of the observed output vector in this output space. At each level, multiple samples of the model parameter vector are generated by a component-wise Metropolis algorithm so that the predicted output corresponding to each parameter value falls in the current data-approximating region. Theoretically, if continued to the limit, the sequence of data-approximating regions would converge on to the observed output vector and the approximate posterior distributions, which are conditional on the data-approximation region, would become exact, but this is not practically feasible. In this paper we study the performance of the ABC-SubSim algorithm for Bayesian updating of the parameters of dynamical systems using a general hierarchical state-space model. We note that the ABC methodology gives an approximate posterior distribution that actually corresponds to an exact posterior where a uniformly distributed combined measurement and modeling error is added. We also note that ABC algorithms have a problem with learning the uncertain error variances in a stochastic state-space model and so we treat them as nuisance parameters and analytically integrate them out of the posterior distribution. In addition, the statistical efficiency of the original ABC-SubSim algorithm is improved by developing a novel strategy to regulate the proposal variance for the component-wise Metropolis algorithm at each level. We demonstrate that Self-regulated ABC-SubSim is well suited for Bayesian system identification by first applying it successfully to model updating of a two degree-of-freedom linear structure for three cases: globally, locally and un-identifiable model classes, and then to model updating of a two degree-of-freedom nonlinear structure with Duffing nonlinearities in its interstory force-deflection relationship.

Learning to cooperate without awareness in multiplayer minimal social situations.

PubMed

Colman, Andrew M; Pulford, Briony D; Omtzigt, David; al-Nowaihi, Ali

2010-11-01

Experimental and Monte Carlo methods were used to test theoretical predictions about adaptive learning of cooperative responses without awareness in minimal social situations-games in which the payoffs to players depend not on their own actions but exclusively on the actions of other group members. In Experiment 1, learning occurred slowly over 200 rounds in a dyadic minimal social situation but not in multiplayer groups. In Experiments 2-4, learning occurred rarely in multiplayer groups, even when players were informed that they were interacting strategically and were allowed to communicate with one another but were not aware of the game's payoff structure. Monte Carlo simulation suggested that players approach minimal social situations using a noisy version of the win-stay, lose-shift decision rule, deviating from the deterministic rule less frequently after rewarding than unrewarding rounds. Copyright 2010 Elsevier Inc. All rights reserved.

Bayesian forecasting and uncertainty quantifying of stream flows using Metropolis–Hastings Markov Chain Monte Carlo algorithm

DOE PAGES

Wang, Hongrui; Wang, Cheng; Wang, Ying; ...

2017-04-05

This paper presents a Bayesian approach using Metropolis-Hastings Markov Chain Monte Carlo algorithm and applies this method for daily river flow rate forecast and uncertainty quantification for Zhujiachuan River using data collected from Qiaotoubao Gage Station and other 13 gage stations in Zhujiachuan watershed in China. The proposed method is also compared with the conventional maximum likelihood estimation (MLE) for parameter estimation and quantification of associated uncertainties. While the Bayesian method performs similarly in estimating the mean value of daily flow rate, it performs over the conventional MLE method on uncertainty quantification, providing relatively narrower reliable interval than the MLEmore » confidence interval and thus more precise estimation by using the related information from regional gage stations. As a result, the Bayesian MCMC method might be more favorable in the uncertainty analysis and risk management.« less

Marathon: An Open Source Software Library for the Analysis of Markov-Chain Monte Carlo Algorithms

PubMed Central

Rechner, Steffen; Berger, Annabell

2016-01-01

We present the software library marathon, which is designed to support the analysis of sampling algorithms that are based on the Markov-Chain Monte Carlo principle. The main application of this library is the computation of properties of so-called state graphs, which represent the structure of Markov chains. We demonstrate applications and the usefulness of marathon by investigating the quality of several bounding methods on four well-known Markov chains for sampling perfect matchings and bipartite graphs. In a set of experiments, we compute the total mixing time and several of its bounds for a large number of input instances. We find that the upper bound gained by the famous canonical path method is often several magnitudes larger than the total mixing time and deteriorates with growing input size. In contrast, the spectral bound is found to be a precise approximation of the total mixing time. PMID:26824442

Methods for Monte Carlo simulations of biomacromolecules

PubMed Central

Vitalis, Andreas; Pappu, Rohit V.

2010-01-01

The state-of-the-art for Monte Carlo (MC) simulations of biomacromolecules is reviewed. Available methodologies for sampling conformational equilibria and associations of biomacromolecules in the canonical ensemble, given a continuum description of the solvent environment, are reviewed. Detailed sections are provided dealing with the choice of degrees of freedom, the efficiencies of MC algorithms and algorithmic peculiarities, as well as the optimization of simple movesets. The issue of introducing correlations into elementary MC moves, and the applicability of such methods to simulations of biomacromolecules is discussed. A brief discussion of multicanonical methods and an overview of recent simulation work highlighting the potential of MC methods are also provided. It is argued that MC simulations, while underutilized biomacromolecular simulation community, hold promise for simulations of complex systems and phenomena that span multiple length scales, especially when used in conjunction with implicit solvation models or other coarse graining strategies. PMID:20428473

SALUTE Grid Application using Message-Oriented Middleware

NASA Astrophysics Data System (ADS)

Atanassov, E.; Dimitrov, D. Sl.; Gurov, T.

2009-10-01

Stochastic ALgorithms for Ultra-fast Transport in sEmiconductors (SALUTE) is a grid application developed for solving various computationally intensive problems which describe ultra-fast carrier transport in semiconductors. SALUTE studies memory and quantum effects during the relaxation process due to electronphonon interaction in one-band semiconductors or quantum wires. Formally, SALUTE integrates a set of novel Monte Carlo, quasi-Monte Carlo and hybrid algorithms for solving various computationally intensive problems which describe the femtosecond relaxation process of optically excited carriers in one-band semiconductors or quantum wires. In this paper we present application-specific job submission and reservation management tool named a Job Track Server (JTS). It is developed using Message-Oriented middleware to implement robust, versatile job submission and tracing mechanism, which can be tailored to application specific failover and quality of service requirements. Experience from using the JTS for submission of SALUTE jobs is presented.

System identification and model reduction using modulating function techniques

NASA Technical Reports Server (NTRS)

Shen, Yan

1993-01-01

Weighted least squares (WLS) and adaptive weighted least squares (AWLS) algorithms are initiated for continuous-time system identification using Fourier type modulating function techniques. Two stochastic signal models are examined using the mean square properties of the stochastic calculus: an equation error signal model with white noise residuals, and a more realistic white measurement noise signal model. The covariance matrices in each model are shown to be banded and sparse, and a joint likelihood cost function is developed which links the real and imaginary parts of the modulated quantities. The superior performance of above algorithms is demonstrated by comparing them with the LS/MFT and popular predicting error method (PEM) through 200 Monte Carlo simulations. A model reduction problem is formulated with the AWLS/MFT algorithm, and comparisons are made via six examples with a variety of model reduction techniques, including the well-known balanced realization method. Here the AWLS/MFT algorithm manifests higher accuracy in almost all cases, and exhibits its unique flexibility and versatility. Armed with this model reduction, the AWLS/MFT algorithm is extended into MIMO transfer function system identification problems. The impact due to the discrepancy in bandwidths and gains among subsystem is explored through five examples. Finally, as a comprehensive application, the stability derivatives of the longitudinal and lateral dynamics of an F-18 aircraft are identified using physical flight data provided by NASA. A pole-constrained SIMO and MIMO AWLS/MFT algorithm is devised and analyzed. Monte Carlo simulations illustrate its high-noise rejecting properties. Utilizing the flight data, comparisons among different MFT algorithms are tabulated and the AWLS is found to be strongly favored in almost all facets.

Monte Carlo methods for multidimensional integration for European option pricing

NASA Astrophysics Data System (ADS)

Todorov, V.; Dimov, I. T.

2016-10-01

In this paper, we illustrate examples of highly accurate Monte Carlo and quasi-Monte Carlo methods for multiple integrals related to the evaluation of European style options. The idea is that the value of the option is formulated in terms of the expectation of some random variable; then the average of independent samples of this random variable is used to estimate the value of the option. First we obtain an integral representation for the value of the option using the risk neutral valuation formula. Then with an appropriations change of the constants we obtain a multidimensional integral over the unit hypercube of the corresponding dimensionality. Then we compare a specific type of lattice rules over one of the best low discrepancy sequence of Sobol for numerical integration. Quasi-Monte Carlo methods are compared with Adaptive and Crude Monte Carlo techniques for solving the problem. The four approaches are completely different thus it is a question of interest to know which one of them outperforms the other for evaluation multidimensional integrals in finance. Some of the advantages and disadvantages of the developed algorithms are discussed.

On adaptive learning rate that guarantees convergence in feedforward networks.

PubMed

Behera, Laxmidhar; Kumar, Swagat; Patnaik, Awhan

2006-09-01

This paper investigates new learning algorithms (LF I and LF II) based on Lyapunov function for the training of feedforward neural networks. It is observed that such algorithms have interesting parallel with the popular backpropagation (BP) algorithm where the fixed learning rate is replaced by an adaptive learning rate computed using convergence theorem based on Lyapunov stability theory. LF II, a modified version of LF I, has been introduced with an aim to avoid local minima. This modification also helps in improving the convergence speed in some cases. Conditions for achieving global minimum for these kind of algorithms have been studied in detail. The performances of the proposed algorithms are compared with BP algorithm and extended Kalman filtering (EKF) on three bench-mark function approximation problems: XOR, 3-bit parity, and 8-3 encoder. The comparisons are made in terms of number of learning iterations and computational time required for convergence. It is found that the proposed algorithms (LF I and II) are much faster in convergence than other two algorithms to attain same accuracy. Finally, the comparison is made on a complex two-dimensional (2-D) Gabor function and effect of adaptive learning rate for faster convergence is verified. In a nutshell, the investigations made in this paper help us better understand the learning procedure of feedforward neural networks in terms of adaptive learning rate, convergence speed, and local minima.

Optimal signal constellation design for ultra-high-speed optical transport in the presence of nonlinear phase noise.

PubMed

Liu, Tao; Djordjevic, Ivan B

2014-12-29

In this paper, we first describe an optimal signal constellation design algorithm suitable for the coherent optical channels dominated by the linear phase noise. Then, we modify this algorithm to be suitable for the nonlinear phase noise dominated channels. In optimization procedure, the proposed algorithm uses the cumulative log-likelihood function instead of the Euclidian distance. Further, an LDPC coded modulation scheme is proposed to be used in combination with signal constellations obtained by proposed algorithm. Monte Carlo simulations indicate that the LDPC-coded modulation schemes employing the new constellation sets, obtained by our new signal constellation design algorithm, outperform corresponding QAM constellations significantly in terms of transmission distance and have better nonlinearity tolerance.

Selection method of terrain matching area for TERCOM algorithm

NASA Astrophysics Data System (ADS)

Zhang, Qieqie; Zhao, Long

2017-10-01

The performance of terrain aided navigation is closely related to the selection of terrain matching area. The different matching algorithms have different adaptability to terrain. This paper mainly studies the adaptability to terrain of TERCOM algorithm, analyze the relation between terrain feature and terrain characteristic parameters by qualitative and quantitative methods, and then research the relation between matching probability and terrain characteristic parameters by the Monte Carlo method. After that, we propose a selection method of terrain matching area for TERCOM algorithm, and verify the method correctness with real terrain data by simulation experiment. Experimental results show that the matching area obtained by the method in this paper has the good navigation performance and the matching probability of TERCOM algorithm is great than 90%

Challenges in the Verification of Reinforcement Learning Algorithms

NASA Technical Reports Server (NTRS)

Van Wesel, Perry; Goodloe, Alwyn E.

2017-01-01

Machine learning (ML) is increasingly being applied to a wide array of domains from search engines to autonomous vehicles. These algorithms, however, are notoriously complex and hard to verify. This work looks at the assumptions underlying machine learning algorithms as well as some of the challenges in trying to verify ML algorithms. Furthermore, we focus on the specific challenges of verifying reinforcement learning algorithms. These are highlighted using a specific example. Ultimately, we do not offer a solution to the complex problem of ML verification, but point out possible approaches for verification and interesting research opportunities.

Semi-supervised learning via regularized boosting working on multiple semi-supervised assumptions.

PubMed

Chen, Ke; Wang, Shihai

2011-01-01

Semi-supervised learning concerns the problem of learning in the presence of labeled and unlabeled data. Several boosting algorithms have been extended to semi-supervised learning with various strategies. To our knowledge, however, none of them takes all three semi-supervised assumptions, i.e., smoothness, cluster, and manifold assumptions, together into account during boosting learning. In this paper, we propose a novel cost functional consisting of the margin cost on labeled data and the regularization penalty on unlabeled data based on three fundamental semi-supervised assumptions. Thus, minimizing our proposed cost functional with a greedy yet stagewise functional optimization procedure leads to a generic boosting framework for semi-supervised learning. Extensive experiments demonstrate that our algorithm yields favorite results for benchmark and real-world classification tasks in comparison to state-of-the-art semi-supervised learning algorithms, including newly developed boosting algorithms. Finally, we discuss relevant issues and relate our algorithm to the previous work.

Communication: An efficient and accurate perturbative correction to initiator full configuration interaction quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Blunt, Nick S.

2018-06-01

We present a perturbative correction within initiator full configuration interaction quantum Monte Carlo (i-FCIQMC). In the existing i-FCIQMC algorithm, a significant number of spawned walkers are discarded due to the initiator criteria. Here we show that these discarded walkers have a form that allows the calculation of a second-order Epstein-Nesbet correction, which may be accumulated in a trivial and inexpensive manner, yet substantially improves i-FCIQMC results. The correction is applied to the Hubbard model and the uniform electron gas and molecular systems.

Quantum Monte Carlo with very large multideterminant wavefunctions.

PubMed

Scemama, Anthony; Applencourt, Thomas; Giner, Emmanuel; Caffarel, Michel

2016-07-01

An algorithm to compute efficiently the first two derivatives of (very) large multideterminant wavefunctions for quantum Monte Carlo calculations is presented. The calculation of determinants and their derivatives is performed using the Sherman-Morrison formula for updating the inverse Slater matrix. An improved implementation based on the reduction of the number of column substitutions and on a very efficient implementation of the calculation of the scalar products involved is presented. It is emphasized that multideterminant expansions contain in general a large number of identical spin-specific determinants: for typical configuration interaction-type wavefunctions the number of unique spin-specific determinants Ndetσ ( σ=↑,↓) with a non-negligible weight in the expansion is of order O(Ndet). We show that a careful implementation of the calculation of the Ndet -dependent contributions can make this step negligible enough so that in practice the algorithm scales as the total number of unique spin-specific determinants,  Ndet↑+Ndet↓, over a wide range of total number of determinants (here, Ndet up to about one million), thus greatly reducing the total computational cost. Finally, a new truncation scheme for the multideterminant expansion is proposed so that larger expansions can be considered without increasing the computational time. The algorithm is illustrated with all-electron fixed-node diffusion Monte Carlo calculations of the total energy of the chlorine atom. Calculations using a trial wavefunction including about 750,000 determinants with a computational increase of ∼400 compared to a single-determinant calculation are shown to be feasible. © 2016 Wiley Periodicals, Inc. © 2016 Wiley Periodicals, Inc.

Apparatus and method for tracking a molecule or particle in three dimensions

DOEpatents

Werner, James H [Los Alamos, NM; Goodwin, Peter M [Los Alamos, NM; Lessard, Guillaume [Santa Fe, NM

2009-03-03

An apparatus and method were used to track the movement of fluorescent particles in three dimensions. Control software was used with the apparatus to implement a tracking algorithm for tracking the motion of the individual particles in glycerol/water mixtures. Monte Carlo simulations suggest that the tracking algorithms in combination with the apparatus may be used for tracking the motion of single fluorescent or fluorescently labeled biomolecules in three dimensions.

Efficient Learning Algorithms with Limited Information

ERIC Educational Resources Information Center

De, Anindya

2013-01-01

The thesis explores efficient learning algorithms in settings which are more restrictive than the PAC model of learning (Valiant) in one of the following two senses: (i) The learning algorithm has a very weak access to the unknown function, as in, it does not get labeled samples for the unknown function (ii) The error guarantee required from the…

Compressive sensing of electrocardiogram signals by promoting sparsity on the second-order difference and by using dictionary learning.

PubMed

Pant, Jeevan K; Krishnan, Sridhar

2014-04-01

A new algorithm for the reconstruction of electrocardiogram (ECG) signals and a dictionary learning algorithm for the enhancement of its reconstruction performance for a class of signals are proposed. The signal reconstruction algorithm is based on minimizing the lp pseudo-norm of the second-order difference, called as the lp(2d) pseudo-norm, of the signal. The optimization involved is carried out using a sequential conjugate-gradient algorithm. The dictionary learning algorithm uses an iterative procedure wherein a signal reconstruction and a dictionary update steps are repeated until a convergence criterion is satisfied. The signal reconstruction step is implemented by using the proposed signal reconstruction algorithm and the dictionary update step is implemented by using the linear least-squares method. Extensive simulation results demonstrate that the proposed algorithm yields improved reconstruction performance for temporally correlated ECG signals relative to the state-of-the-art lp(1d)-regularized least-squares and Bayesian learning based algorithms. Also for a known class of signals, the reconstruction performance of the proposed algorithm can be improved by applying it in conjunction with a dictionary obtained using the proposed dictionary learning algorithm.

The energy-dependent electron loss model: backscattering and application to heterogeneous slab media.

PubMed

Lee, Tae Kyu; Sandison, George A

2003-01-21

Electron backscattering has been incorporated into the energy-dependent electron loss (EL) model and the resulting algorithm is applied to predict dose deposition in slab heterogeneous media. This algorithm utilizes a reflection coefficient from the interface that is computed on the basis of Goudsmit-Saunderson theory and an average energy for the backscattered electrons based on Everhart's theory. Predictions of dose deposition in slab heterogeneous media are compared to the Monte Carlo based dose planning method (DPM) and a numerical discrete ordinates method (DOM). The slab media studied comprised water/Pb, water/Al, water/bone, water/bone/water, and water/lung/water, and incident electron beam energies of 10 MeV and 18 MeV. The predicted dose enhancement due to backscattering is accurate to within 3% of dose maximum even for lead as the backscattering medium. Dose discrepancies at large depths beyond the interface were as high as 5% of dose maximum and we speculate that this error may be attributed to the EL model assuming a Gaussian energy distribution for the electrons at depth. The computational cost is low compared to Monte Carlo simulations making the EL model attractive as a fast dose engine for dose optimization algorithms. The predictive power of the algorithm demonstrates that the small angle scattering restriction on the EL model can be overcome while retaining dose calculation accuracy and requiring only one free variable, chi, in the algorithm to be determined in advance of calculation.

A hybrid multiscale Monte Carlo algorithm (HyMSMC) to cope with disparity in time scales and species populations in intracellular networks.

PubMed

Samant, Asawari; Ogunnaike, Babatunde A; Vlachos, Dionisios G

2007-05-24

The fundamental role that intrinsic stochasticity plays in cellular functions has been shown via numerous computational and experimental studies. In the face of such evidence, it is important that intracellular networks are simulated with stochastic algorithms that can capture molecular fluctuations. However, separation of time scales and disparity in species population, two common features of intracellular networks, make stochastic simulation of such networks computationally prohibitive. While recent work has addressed each of these challenges separately, a generic algorithm that can simultaneously tackle disparity in time scales and population scales in stochastic systems is currently lacking. In this paper, we propose the hybrid, multiscale Monte Carlo (HyMSMC) method that fills in this void. The proposed HyMSMC method blends stochastic singular perturbation concepts, to deal with potential stiffness, with a hybrid of exact and coarse-grained stochastic algorithms, to cope with separation in population sizes. In addition, we introduce the computational singular perturbation (CSP) method as a means of systematically partitioning fast and slow networks and computing relaxation times for convergence. We also propose a new criteria of convergence of fast networks to stochastic low-dimensional manifolds, which further accelerates the algorithm. We use several prototype and biological examples, including a gene expression model displaying bistability, to demonstrate the efficiency, accuracy and applicability of the HyMSMC method. Bistable models serve as stringent tests for the success of multiscale MC methods and illustrate limitations of some literature methods.

The energy-dependent electron loss model: backscattering and application to heterogeneous slab media

NASA Astrophysics Data System (ADS)

Lee, Tae Kyu; Sandison, George A.

2003-01-01

Electron backscattering has been incorporated into the energy-dependent electron loss (EL) model and the resulting algorithm is applied to predict dose deposition in slab heterogeneous media. This algorithm utilizes a reflection coefficient from the interface that is computed on the basis of Goudsmit-Saunderson theory and an average energy for the backscattered electrons based on Everhart's theory. Predictions of dose deposition in slab heterogeneous media are compared to the Monte Carlo based dose planning method (DPM) and a numerical discrete ordinates method (DOM). The slab media studied comprised water/Pb, water/Al, water/bone, water/bone/water, and water/lung/water, and incident electron beam energies of 10 MeV and 18 MeV. The predicted dose enhancement due to backscattering is accurate to within 3% of dose maximum even for lead as the backscattering medium. Dose discrepancies at large depths beyond the interface were as high as 5% of dose maximum and we speculate that this error may be attributed to the EL model assuming a Gaussian energy distribution for the electrons at depth. The computational cost is low compared to Monte Carlo simulations making the EL model attractive as a fast dose engine for dose optimization algorithms. The predictive power of the algorithm demonstrates that the small angle scattering restriction on the EL model can be overcome while retaining dose calculation accuracy and requiring only one free variable, χ, in the algorithm to be determined in advance of calculation.

Recognition of strong earthquake-prone areas with a single learning class

NASA Astrophysics Data System (ADS)

Gvishiani, A. D.; Agayan, S. M.; Dzeboev, B. A.; Belov, I. O.

2017-05-01

This article presents a new Barrier recognition algorithm with learning, designed for recognition of earthquake-prone areas. In comparison to the Crust (Kora) algorithm, used by the classical EPA approach, the Barrier algorithm proceeds with learning just on one "pure" high-seismic class. The new algorithm operates in the space of absolute values of the geological-geophysical parameters of the objects. The algorithm is used for recognition of earthquake-prone areas with M ≥ 6.0 in the Caucasus region. Comparative analysis of the Crust and Barrier algorithms justifies their productive coherence.

Reconstruction of 2D PET data with Monte Carlo generated system matrix for generalized natural pixels

NASA Astrophysics Data System (ADS)

Vandenberghe, Stefaan; Staelens, Steven; Byrne, Charles L.; Soares, Edward J.; Lemahieu, Ignace; Glick, Stephen J.

2006-06-01

In discrete detector PET, natural pixels are image basis functions calculated from the response of detector pairs. By using reconstruction with natural pixel basis functions, the discretization of the object into a predefined grid can be avoided. Here, we propose to use generalized natural pixel reconstruction. Using this approach, the basis functions are not the detector sensitivity functions as in the natural pixel case but uniform parallel strips. The backprojection of the strip coefficients results in the reconstructed image. This paper proposes an easy and efficient way to generate the matrix M directly by Monte Carlo simulation. Elements of the generalized natural pixel system matrix are formed by calculating the intersection of a parallel strip with the detector sensitivity function. These generalized natural pixels are easier to use than conventional natural pixels because the final step from solution to a square pixel representation is done by simple backprojection. Due to rotational symmetry in the PET scanner, the matrix M is block circulant and only the first blockrow needs to be stored. Data were generated using a fast Monte Carlo simulator using ray tracing. The proposed method was compared to a listmode MLEM algorithm, which used ray tracing for doing forward and backprojection. Comparison of the algorithms with different phantoms showed that an improved resolution can be obtained using generalized natural pixel reconstruction with accurate system modelling. In addition, it was noted that for the same resolution a lower noise level is present in this reconstruction. A numerical observer study showed the proposed method exhibited increased performance as compared to a standard listmode EM algorithm. In another study, more realistic data were generated using the GATE Monte Carlo simulator. For these data, a more uniform contrast recovery and a better contrast-to-noise performance were observed. It was observed that major improvements in contrast recovery were obtained with MLEM when the correct system matrix was used instead of simple ray tracing. The correct modelling was the major cause of improved contrast for the same background noise. Less important factors were the choice of the algorithm (MLEM performed better than ART) and the basis functions (generalized natural pixels gave better results than pixels).

Bare-Bones Teaching-Learning-Based Optimization

PubMed Central

Zou, Feng; Wang, Lei; Hei, Xinhong; Chen, Debao; Jiang, Qiaoyong; Li, Hongye

2014-01-01

Teaching-learning-based optimization (TLBO) algorithm which simulates the teaching-learning process of the class room is one of the recently proposed swarm intelligent (SI) algorithms. In this paper, a new TLBO variant called bare-bones teaching-learning-based optimization (BBTLBO) is presented to solve the global optimization problems. In this method, each learner of teacher phase employs an interactive learning strategy, which is the hybridization of the learning strategy of teacher phase in the standard TLBO and Gaussian sampling learning based on neighborhood search, and each learner of learner phase employs the learning strategy of learner phase in the standard TLBO or the new neighborhood search strategy. To verify the performance of our approaches, 20 benchmark functions and two real-world problems are utilized. Conducted experiments can been observed that the BBTLBO performs significantly better than, or at least comparable to, TLBO and some existing bare-bones algorithms. The results indicate that the proposed algorithm is competitive to some other optimization algorithms. PMID:25013844

Bare-bones teaching-learning-based optimization.

PubMed

Zou, Feng; Wang, Lei; Hei, Xinhong; Chen, Debao; Jiang, Qiaoyong; Li, Hongye

2014-01-01

Teaching-learning-based optimization (TLBO) algorithm which simulates the teaching-learning process of the class room is one of the recently proposed swarm intelligent (SI) algorithms. In this paper, a new TLBO variant called bare-bones teaching-learning-based optimization (BBTLBO) is presented to solve the global optimization problems. In this method, each learner of teacher phase employs an interactive learning strategy, which is the hybridization of the learning strategy of teacher phase in the standard TLBO and Gaussian sampling learning based on neighborhood search, and each learner of learner phase employs the learning strategy of learner phase in the standard TLBO or the new neighborhood search strategy. To verify the performance of our approaches, 20 benchmark functions and two real-world problems are utilized. Conducted experiments can been observed that the BBTLBO performs significantly better than, or at least comparable to, TLBO and some existing bare-bones algorithms. The results indicate that the proposed algorithm is competitive to some other optimization algorithms.

Comparison of space radiation calculations for deterministic and Monte Carlo transport codes

NASA Astrophysics Data System (ADS)

Lin, Zi-Wei; Adams, James; Barghouty, Abdulnasser; Randeniya, Sharmalee; Tripathi, Ram; Watts, John; Yepes, Pablo

For space radiation protection of astronauts or electronic equipments, it is necessary to develop and use accurate radiation transport codes. Radiation transport codes include deterministic codes, such as HZETRN from NASA and UPROP from the Naval Research Laboratory, and Monte Carlo codes such as FLUKA, the Geant4 toolkit and HETC-HEDS. The deterministic codes and Monte Carlo codes complement each other in that deterministic codes are very fast while Monte Carlo codes are more elaborate. Therefore it is important to investigate how well the results of deterministic codes compare with those of Monte Carlo transport codes and where they differ. In this study we evaluate these different codes in their space radiation applications by comparing their output results in the same given space radiation environments, shielding geometry and material. Typical space radiation environments such as the 1977 solar minimum galactic cosmic ray environment are used as the well-defined input, and simple geometries made of aluminum, water and/or polyethylene are used to represent the shielding material. We then compare various outputs of these codes, such as the dose-depth curves and the flux spectra of different fragments and other secondary particles. These comparisons enable us to learn more about the main differences between these space radiation transport codes. At the same time, they help us to learn the qualitative and quantitative features that these transport codes have in common.

Monte Carlo simulations of time-of-flight PET with double-ended readout: calibration, coincidence resolving times and statistical lower bounds

PubMed Central

Derenzo, Stephen E

2017-01-01

This paper demonstrates through Monte Carlo simulations that a practical positron emission tomograph with (1) deep scintillators for efficient detection, (2) double-ended readout for depth-of-interaction information, (3) fixed-level analog triggering, and (4) accurate calibration and timing data corrections can achieve a coincidence resolving time (CRT) that is not far above the statistical lower bound. One Monte Carlo algorithm simulates a calibration procedure that uses data from a positron point source. Annihilation events with an interaction near the entrance surface of one scintillator are selected, and data from the two photodetectors on the other scintillator provide depth-dependent timing corrections. Another Monte Carlo algorithm simulates normal operation using these corrections and determines the CRT. A third Monte Carlo algorithm determines the CRT statistical lower bound by generating a series of random interaction depths, and for each interaction a set of random photoelectron times for each of the two photodetectors. The most likely interaction times are determined by shifting the depth-dependent probability density function to maximize the joint likelihood for all the photoelectron times in each set. Example calculations are tabulated for different numbers of photoelectrons and photodetector time jitters for three 3 × 3 × 30 mm3 scintillators: Lu2SiO5:Ce,Ca (LSO), LaBr3:Ce, and a hypothetical ultra-fast scintillator. To isolate the factors that depend on the scintillator length and the ability to estimate the DOI, CRT values are tabulated for perfect scintillator-photodetectors. For LSO with 4000 photoelectrons and single photoelectron time jitter of the photodetector J = 0.2 ns (FWHM), the CRT value using the statistically weighted average of corrected trigger times is 0.098 ns FWHM and the statistical lower bound is 0.091 ns FWHM. For LaBr3:Ce with 8000 photoelectrons and J = 0.2 ns FWHM, the CRT values are 0.070 and 0.063 ns FWHM, respectively. For the ultra-fast scintillator with 1 ns decay time, 4000 photoelectrons, and J = 0.2 ns FWHM, the CRT values are 0.021 and 0.017 ns FWHM, respectively. The examples also show that calibration and correction for depth-dependent variations in pulse height and in annihilation and optical photon transit times are necessary to achieve these CRT values. PMID:28327464

Monte Carlo simulations of time-of-flight PET with double-ended readout: calibration, coincidence resolving times and statistical lower bounds

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

Derenzo, Stephen E.

Here, this paper demonstrates through Monte Carlo simulations that a practical positron emission tomograph with (1) deep scintillators for efficient detection, (2) double-ended readout for depth-of-interaction information, (3) fixed-level analog triggering, and (4) accurate calibration and timing data corrections can achieve a coincidence resolving time (CRT) that is not far above the statistical lower bound. One Monte Carlo algorithm simulates a calibration procedure that uses data from a positron point source. Annihilation events with an interaction near the entrance surface of one scintillator are selected, and data from the two photodetectors on the other scintillator provide depth-dependent timing corrections. Anothermore » Monte Carlo algorithm simulates normal operation using these corrections and determines the CRT. A third Monte Carlo algorithm determines the CRT statistical lower bound by generating a series of random interaction depths, and for each interaction a set of random photoelectron times for each of the two photodetectors. The most likely interaction times are determined by shifting the depth-dependent probability density function to maximize the joint likelihood for all the photoelectron times in each set. Example calculations are tabulated for different numbers of photoelectrons and photodetector time jitters for three 3 × 3 × 30 mm 3 scintillators: Lu 2SiO 5 :Ce,Ca (LSO), LaBr 3:Ce, and a hypothetical ultra-fast scintillator. To isolate the factors that depend on the scintillator length and the ability to estimate the DOI, CRT values are tabulated for perfect scintillator-photodetectors. For LSO with 4000 photoelectrons and single photoelectron time jitter of the photodetector J = 0.2 ns (FWHM), the CRT value using the statistically weighted average of corrected trigger times is 0.098 ns FWHM and the statistical lower bound is 0.091 ns FWHM. For LaBr 3:Ce with 8000 photoelectrons and J = 0.2 ns FWHM, the CRT values are 0.070 and 0.063 ns FWHM, respectively. For the ultra-fast scintillator with 1 ns decay time, 4000 photoelectrons, and J = 0.2 ns FWHM, the CRT values are 0.021 and 0.017 ns FWHM, respectively. Lastly, the examples also show that calibration and correction for depth-dependent variations in pulse height and in annihilation and optical photon transit times are necessary to achieve these CRT values.« less

Monte Carlo simulations of time-of-flight PET with double-ended readout: calibration, coincidence resolving times and statistical lower bounds

DOE PAGES

Derenzo, Stephen E.

2017-04-11

Here, this paper demonstrates through Monte Carlo simulations that a practical positron emission tomograph with (1) deep scintillators for efficient detection, (2) double-ended readout for depth-of-interaction information, (3) fixed-level analog triggering, and (4) accurate calibration and timing data corrections can achieve a coincidence resolving time (CRT) that is not far above the statistical lower bound. One Monte Carlo algorithm simulates a calibration procedure that uses data from a positron point source. Annihilation events with an interaction near the entrance surface of one scintillator are selected, and data from the two photodetectors on the other scintillator provide depth-dependent timing corrections. Anothermore » Monte Carlo algorithm simulates normal operation using these corrections and determines the CRT. A third Monte Carlo algorithm determines the CRT statistical lower bound by generating a series of random interaction depths, and for each interaction a set of random photoelectron times for each of the two photodetectors. The most likely interaction times are determined by shifting the depth-dependent probability density function to maximize the joint likelihood for all the photoelectron times in each set. Example calculations are tabulated for different numbers of photoelectrons and photodetector time jitters for three 3 × 3 × 30 mm 3 scintillators: Lu 2SiO 5 :Ce,Ca (LSO), LaBr 3:Ce, and a hypothetical ultra-fast scintillator. To isolate the factors that depend on the scintillator length and the ability to estimate the DOI, CRT values are tabulated for perfect scintillator-photodetectors. For LSO with 4000 photoelectrons and single photoelectron time jitter of the photodetector J = 0.2 ns (FWHM), the CRT value using the statistically weighted average of corrected trigger times is 0.098 ns FWHM and the statistical lower bound is 0.091 ns FWHM. For LaBr 3:Ce with 8000 photoelectrons and J = 0.2 ns FWHM, the CRT values are 0.070 and 0.063 ns FWHM, respectively. For the ultra-fast scintillator with 1 ns decay time, 4000 photoelectrons, and J = 0.2 ns FWHM, the CRT values are 0.021 and 0.017 ns FWHM, respectively. Lastly, the examples also show that calibration and correction for depth-dependent variations in pulse height and in annihilation and optical photon transit times are necessary to achieve these CRT values.« less

Monte Carlo PDF method for turbulent reacting flow in a jet-stirred reactor

NASA Astrophysics Data System (ADS)

Roekaerts, D.

1992-01-01

A stochastic algorithm for the solution of the modeled scalar probability density function (PDF) transport equation for single-phase turbulent reacting flow is described. Cylindrical symmetry is assumed. The PDF is represented by ensembles of N representative values of the thermochemical variables in each cell of a nonuniform finite-difference grid and operations on these elements representing convection, diffusion, mixing and reaction are derived. A simplified model and solution algorithm which neglects the influence of turbulent fluctuations on mean reaction rates is also described. Both algorithms are applied to a selectivity problem in a real reactor.

Online Feature Transformation Learning for Cross-Domain Object Category Recognition.

PubMed

Zhang, Xuesong; Zhuang, Yan; Wang, Wei; Pedrycz, Witold

2017-06-09

In this paper, we introduce a new research problem termed online feature transformation learning in the context of multiclass object category recognition. The learning of a feature transformation is viewed as learning a global similarity metric function in an online manner. We first consider the problem of online learning a feature transformation matrix expressed in the original feature space and propose an online passive aggressive feature transformation algorithm. Then these original features are mapped to kernel space and an online single kernel feature transformation (OSKFT) algorithm is developed to learn a nonlinear feature transformation. Based on the OSKFT and the existing Hedge algorithm, a novel online multiple kernel feature transformation algorithm is also proposed, which can further improve the performance of online feature transformation learning in large-scale application. The classifier is trained with k nearest neighbor algorithm together with the learned similarity metric function. Finally, we experimentally examined the effect of setting different parameter values in the proposed algorithms and evaluate the model performance on several multiclass object recognition data sets. The experimental results demonstrate the validity and good performance of our methods on cross-domain and multiclass object recognition application.

Controlling false-negative errors in microarray differential expression analysis: a PRIM approach.

PubMed

Cole, Steve W; Galic, Zoran; Zack, Jerome A

2003-09-22

Theoretical considerations suggest that current microarray screening algorithms may fail to detect many true differences in gene expression (Type II analytic errors). We assessed 'false negative' error rates in differential expression analyses by conventional linear statistical models (e.g. t-test), microarray-adapted variants (e.g. SAM, Cyber-T), and a novel strategy based on hold-out cross-validation. The latter approach employs the machine-learning algorithm Patient Rule Induction Method (PRIM) to infer minimum thresholds for reliable change in gene expression from Boolean conjunctions of fold-induction and raw fluorescence measurements. Monte Carlo analyses based on four empirical data sets show that conventional statistical models and their microarray-adapted variants overlook more than 50% of genes showing significant up-regulation. Conjoint PRIM prediction rules recover approximately twice as many differentially expressed transcripts while maintaining strong control over false-positive (Type I) errors. As a result, experimental replication rates increase and total analytic error rates decline. RT-PCR studies confirm that gene inductions detected by PRIM but overlooked by other methods represent true changes in mRNA levels. PRIM-based conjoint inference rules thus represent an improved strategy for high-sensitivity screening of DNA microarrays. Freestanding JAVA application at http://microarray.crump.ucla.edu/focus

Coupled particle-in-cell and Monte Carlo transport modeling of intense radiographic sources

NASA Astrophysics Data System (ADS)

Rose, D. V.; Welch, D. R.; Oliver, B. V.; Clark, R. E.; Johnson, D. L.; Maenchen, J. E.; Menge, P. R.; Olson, C. L.; Rovang, D. C.

2002-03-01

Dose-rate calculations for intense electron-beam diodes using particle-in-cell (PIC) simulations along with Monte Carlo electron/photon transport calculations are presented. The electromagnetic PIC simulations are used to model the dynamic operation of the rod-pinch and immersed-B diodes. These simulations include algorithms for tracking electron scattering and energy loss in dense materials. The positions and momenta of photons created in these materials are recorded and separate Monte Carlo calculations are used to transport the photons to determine the dose in far-field detectors. These combined calculations are used to determine radiographer equations (dose scaling as a function of diode current and voltage) that are compared directly with measured dose rates obtained on the SABRE generator at Sandia National Laboratories.

A linear recurrent kernel online learning algorithm with sparse updates.

PubMed

Fan, Haijin; Song, Qing

2014-02-01

In this paper, we propose a recurrent kernel algorithm with selectively sparse updates for online learning. The algorithm introduces a linear recurrent term in the estimation of the current output. This makes the past information reusable for updating of the algorithm in the form of a recurrent gradient term. To ensure that the reuse of this recurrent gradient indeed accelerates the convergence speed, a novel hybrid recurrent training is proposed to switch on or off learning the recurrent information according to the magnitude of the current training error. Furthermore, the algorithm includes a data-dependent adaptive learning rate which can provide guaranteed system weight convergence at each training iteration. The learning rate is set as zero when the training violates the derived convergence conditions, which makes the algorithm updating process sparse. Theoretical analyses of the weight convergence are presented and experimental results show the good performance of the proposed algorithm in terms of convergence speed and estimation accuracy. Copyright © 2013 Elsevier Ltd. All rights reserved.

Kernel Temporal Differences for Neural Decoding

PubMed Central

Bae, Jihye; Sanchez Giraldo, Luis G.; Pohlmeyer, Eric A.; Francis, Joseph T.; Sanchez, Justin C.; Príncipe, José C.

2015-01-01

We study the feasibility and capability of the kernel temporal difference (KTD)(λ) algorithm for neural decoding. KTD(λ) is an online, kernel-based learning algorithm, which has been introduced to estimate value functions in reinforcement learning. This algorithm combines kernel-based representations with the temporal difference approach to learning. One of our key observations is that by using strictly positive definite kernels, algorithm's convergence can be guaranteed for policy evaluation. The algorithm's nonlinear functional approximation capabilities are shown in both simulations of policy evaluation and neural decoding problems (policy improvement). KTD can handle high-dimensional neural states containing spatial-temporal information at a reasonable computational complexity allowing real-time applications. When the algorithm seeks a proper mapping between a monkey's neural states and desired positions of a computer cursor or a robot arm, in both open-loop and closed-loop experiments, it can effectively learn the neural state to action mapping. Finally, a visualization of the coadaptation process between the decoder and the subject shows the algorithm's capabilities in reinforcement learning brain machine interfaces. PMID:25866504

TH-A-19A-06: Site-Specific Comparison of Analytical and Monte Carlo Based Dose Calculations

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

Schuemann, J; Grassberger, C; Paganetti, H

2014-06-15

Purpose: To investigate the impact of complex patient geometries on the capability of analytical dose calculation algorithms to accurately predict dose distributions and to verify currently used uncertainty margins in proton therapy. Methods: Dose distributions predicted by an analytical pencilbeam algorithm were compared with Monte Carlo simulations (MCS) using TOPAS. 79 complete patient treatment plans were investigated for 7 disease sites (liver, prostate, breast, medulloblastoma spine and whole brain, lung and head and neck). A total of 508 individual passively scattered treatment fields were analyzed for field specific properties. Comparisons based on target coverage indices (EUD, D95, D90 and D50)more » were performed. Range differences were estimated for the distal position of the 90% dose level (R90) and the 50% dose level (R50). Two-dimensional distal dose surfaces were calculated and the root mean square differences (RMSD), average range difference (ARD) and average distal dose degradation (ADD), the distance between the distal position of the 80% and 20% dose levels (R80- R20), were analyzed. Results: We found target coverage indices calculated by TOPAS to generally be around 1–2% lower than predicted by the analytical algorithm. Differences in R90 predicted by TOPAS and the planning system can be larger than currently applied range margins in proton therapy for small regions distal to the target volume. We estimate new site-specific range margins (R90) for analytical dose calculations considering total range uncertainties and uncertainties from dose calculation alone based on the RMSD. Our results demonstrate that a reduction of currently used uncertainty margins is feasible for liver, prostate and whole brain fields even without introducing MC dose calculations. Conclusion: Analytical dose calculation algorithms predict dose distributions within clinical limits for more homogeneous patients sites (liver, prostate, whole brain). However, we recommend treatment plan verification using Monte Carlo simulations for patients with complex geometries.« less

Analysis of the penumbra enlargement in lung versus the quality index of photon beams: a methodology to check the dose calculation algorithm.

PubMed

Tsiakalos, Miltiadis F; Theodorou, Kiki; Kappas, Constantin; Zefkili, Sofia; Rosenwold, Jean-Claude

2004-04-01

It is well known that considerable underdosage can occur at the edges of a tumor inside the lung because of the degradation of penumbra due to lack of lateral electronic equilibrium. Although present even at smaller energies, this phenomenon is more pronounced for higher energies. Apart from Monte Carlo calculation, most of the existing Treatment Planning Systems (TPSs) cannot deal at all, or with acceptable accuracy, with this effect. A methodology has been developed for assessing the dose calculation algorithms in the lung region where lateral electronic disequilibrium exists, based on the Quality Index (QI) of the incident beam. A phantom, consisting of layers of polystyrene and lung material, has been irradiated using photon beams of 4, 6, 15, and 20 MV. The cross-plane profiles of each beam for 5x5, 10x10, and 25x10 fields have been measured at the middle of the phantom with the use of films. The penumbra (20%-80%) and fringe (50%-90%) enlargement was measured and the ratio of the widths for the lung to that of polystyrene was defined as the Correction Factor (CF). Monte Carlo calculations in the two phantoms have also been performed for energies of 6, 15, and 20 MV. Five commercial TPS's algorithms were tested for their ability to predict the penumbra and fringe enlargement. A linear relationship has been found between the QI of the beams and the CF of the penumbra and fringe enlargement for all the examined fields. Monte Carlo calculations agree very well (less than 1% difference) with the film measurements. The CF values range between 1.1 for 4 MV (QI 0.620) and 2.28 for 20 MV (QI 0.794). Three of the tested TPS's algorithms could not predict any enlargement at all for all energies and all fields and two of them could predict the penumbra enlargement to some extent. The proposed methodology can help any user or developer to check the accuracy of its algorithm for lung cases, based on a simple phantom geometry and the QI of the incident beam. This check is very important especially when higher energies are used, as the inaccuracies in existing algorithms can lead to an incorrect choice of energy for lung treatment and consequently to a failure in tumor control.

Computer Simulation of Electron Positron Annihilation Processes

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

Chen, y

2003-10-02

With the launching of the Next Linear Collider coming closer and closer, there is a pressing need for physicists to develop a fully-integrated computer simulation of e{sup +}e{sup -} annihilation process at center-of-mass energy of 1TeV. A simulation program acts as the template for future experiments. Either new physics will be discovered, or current theoretical uncertainties will shrink due to more accurate higher-order radiative correction calculations. The existence of an efficient and accurate simulation will help us understand the new data and validate (or veto) some of the theoretical models developed to explain new physics. It should handle well interfacesmore » between different sectors of physics, e.g., interactions happening at parton levels well above the QCD scale which are described by perturbative QCD, and interactions happening at much lower energy scale, which combine partons into hadrons. Also it should achieve competitive speed in real time when the complexity of the simulation increases. This thesis contributes some tools that will be useful for the development of such simulation programs. We begin our study by the development of a new Monte Carlo algorithm intended to perform efficiently in selecting weight-1 events when multiple parameter dimensions are strongly correlated. The algorithm first seeks to model the peaks of the distribution by features, adapting these features to the function using the EM algorithm. The representation of the distribution provided by these features is then improved using the VEGAS algorithm for the Monte Carlo integration. The two strategies mesh neatly into an effective multi-channel adaptive representation. We then present a new algorithm for the simulation of parton shower processes in high energy QCD. We want to find an algorithm which is free of negative weights, produces its output as a set of exclusive events, and whose total rate exactly matches the full Feynman amplitude calculation. Our strategy is to create the whole QCD shower as a tree structure generated by a multiple Poisson process. Working with the whole shower allows us to include correlations between gluon emissions from different sources. QCD destructive interference is controlled by the implementation of ''angular-ordering,'' as in the HERWIG Monte Carlo program. We discuss methods for systematic improvement of the approach to include higher order QCD effects.« less

Photon beam dosimetry with EBT3 film in heterogeneous regions: Application to the evaluation of dose-calculation algorithms

NASA Astrophysics Data System (ADS)

Jung, Hyunuk; Kum, Oyeon; Han, Youngyih; Park, Byungdo; Cheong, Kwang-Ho

2014-12-01

For a better understanding of the accuracy of state-of-the-art-radiation therapies, 2-dimensional dosimetry in a patient-like environment will be helpful. Therefore, the dosimetry of EBT3 films in non-water-equivalent tissues was investigated, and the accuracy of commercially-used dose-calculation algorithms was evaluated with EBT3 measurement. Dose distributions were measured with EBT3 films for an in-house-designed phantom that contained a lung or a bone substitute, i.e., an air cavity (3 × 3 × 3 cm3) or teflon (2 × 2 × 2 cm3 or 3 × 3 × 3 cm3), respectively. The phantom was irradiated with 6-MV X-rays with field sizes of 2 × 2, 3 × 3, and 5 × 5 cm2. The accuracy of EBT3 dosimetry was evaluated by comparing the measured dose with the dose obtained from Monte Carlo (MC) simulations. A dose-to-bone-equivalent material was obtained by multiplying the EBT3 measurements by the stopping power ratio (SPR). The EBT3 measurements were then compared with the predictions from four algorithms: Monte Carlo (MC) in iPlan, acuros XB (AXB), analytical anisotropic algorithm (AAA) in Eclipse, and superposition-convolution (SC) in Pinnacle. For the air cavity, the EBT3 measurements agreed with the MC calculation to within 2% on average. For teflon, the EBT3 measurements differed by 9.297% (±0.9229%) on average from the Monte Carlo calculation before dose conversion, and by 0.717% (±0.6546%) after applying the SPR. The doses calculated by using the MC, AXB, AAA, and SC algorithms for the air cavity differed from the EBT3 measurements on average by 2.174, 2.863, 18.01, and 8.391%, respectively; for teflon, the average differences were 3.447, 4.113, 7.589, and 5.102%. The EBT3 measurements corrected with the SPR agreed with 2% on average both within and beyond the heterogeneities with MC results, thereby indicating that EBT3 dosimetry can be used in heterogeneous media. The MC and the AXB dose calculation algorithms exhibited clinically-acceptable accuracy (<5%) in heterogeneities.

The Wang-Landau Sampling Algorithm

NASA Astrophysics Data System (ADS)

Landau, David P.

2003-03-01

Over the past several decades Monte Carlo simulations[1] have evolved into a powerful tool for the study of wide-ranging problems in statistical/condensed matter physics. Standard methods sample the probability distribution for the states of the system, usually in the canonical ensemble, and enormous improvements have been made in performance through the implementation of novel algorithms. Nonetheless, difficulties arise near phase transitions, either due to critical slowing down near 2nd order transitions or to metastability near 1st order transitions, thus limiting the applicability of the method. We shall describe a new and different Monte Carlo approach [2] that uses a random walk in energy space to determine the density of states directly. Once the density of states is estimated, all thermodynamic properties can be calculated at all temperatures. This approach can be extended to multi-dimensional parameter spaces and has already found use in classical models of interacting particles including systems with complex energy landscapes, e.g., spin glasses, protein folding models, etc., as well as for quantum models. 1. A Guide to Monte Carlo Simulations in Statistical Physics, D. P. Landau and K. Binder (Cambridge U. Press, Cambridge, 2000). 2. Fugao Wang and D. P. Landau, Phys. Rev. Lett. 86, 2050 (2001); Phys. Rev. E64, 056101-1 (2001).

Accuracy of tree diameter estimation from terrestrial laser scanning by circle-fitting methods

NASA Astrophysics Data System (ADS)

Koreň, Milan; Mokroš, Martin; Bucha, Tomáš

2017-12-01

This study compares the accuracies of diameter at breast height (DBH) estimations by three initial (minimum bounding box, centroid, and maximum distance) and two refining (Monte Carlo and optimal circle) circle-fitting methods The circle-fitting algorithms were evaluated in multi-scan mode and a simulated single-scan mode on 157 European beech trees (Fagus sylvatica L.). DBH measured by a calliper was used as reference data. Most of the studied circle-fitting algorithms significantly underestimated the mean DBH in both scanning modes. Only the Monte Carlo method in the single-scan mode significantly overestimated the mean DBH. The centroid method proved to be the least suitable and showed significantly different results from the other circle-fitting methods in both scanning modes. In multi-scan mode, the accuracy of the minimum bounding box method was not significantly different from the accuracies of the refining methods The accuracy of the maximum distance method was significantly different from the accuracies of the refining methods in both scanning modes. The accuracy of the Monte Carlo method was significantly different from the accuracy of the optimal circle method in only single-scan mode. The optimal circle method proved to be the most accurate circle-fitting method for DBH estimation from point clouds in both scanning modes.

Comparing multiple turbulence restoration algorithms performance on noisy anisoplanatic imagery

NASA Astrophysics Data System (ADS)

Rucci, Michael A.; Hardie, Russell C.; Dapore, Alexander J.

2017-05-01

In this paper, we compare the performance of multiple turbulence mitigation algorithms to restore imagery degraded by atmospheric turbulence and camera noise. In order to quantify and compare algorithm performance, imaging scenes were simulated by applying noise and varying levels of turbulence. For the simulation, a Monte-Carlo wave optics approach is used to simulate the spatially and temporally varying turbulence in an image sequence. A Poisson-Gaussian noise mixture model is then used to add noise to the observed turbulence image set. These degraded image sets are processed with three separate restoration algorithms: Lucky Look imaging, bispectral speckle imaging, and a block matching method with restoration filter. These algorithms were chosen because they incorporate different approaches and processing techniques. The results quantitatively show how well the algorithms are able to restore the simulated degraded imagery.

A sampling algorithm for segregation analysis

PubMed Central

Tier, Bruce; Henshall, John

2001-01-01

Methods for detecting Quantitative Trait Loci (QTL) without markers have generally used iterative peeling algorithms for determining genotype probabilities. These algorithms have considerable shortcomings in complex pedigrees. A Monte Carlo Markov chain (MCMC) method which samples the pedigree of the whole population jointly is described. Simultaneous sampling of the pedigree was achieved by sampling descent graphs using the Metropolis-Hastings algorithm. A descent graph describes the inheritance state of each allele and provides pedigrees guaranteed to be consistent with Mendelian sampling. Sampling descent graphs overcomes most, if not all, of the limitations incurred by iterative peeling algorithms. The algorithm was able to find the QTL in most of the simulated populations. However, when the QTL was not modeled or found then its effect was ascribed to the polygenic component. No QTL were detected when they were not simulated. PMID:11742631

Stan : A Probabilistic Programming Language

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

Carpenter, Bob; Gelman, Andrew; Hoffman, Matthew D.

Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectationmore » propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can also be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.« less

Many-body calculations of low energy eigenstates in magnetic and periodic systems with self healing diffusion Monte Carlo: steps beyond the fixed-phase

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

Reboredo, Fernando A.

The self-healing diffusion Monte Carlo algorithm (SHDMC) [Reboredo, Hood and Kent, Phys. Rev. B {\\bf 79}, 195117 (2009), Reboredo, {\\it ibid.} {\\bf 80}, 125110 (2009)] is extended to study the ground and excited states of magnetic and periodic systems. A recursive optimization algorithm is derived from the time evolution of the mixed probability density. The mixed probability density is given by an ensemble of electronic configurations (walkers) with complex weight. This complex weigh allows the amplitude of the fix-node wave function to move away from the trial wave function phase. This novel approach is both a generalization of SHDMC andmore » the fixed-phase approximation [Ortiz, Ceperley and Martin Phys Rev. Lett. {\\bf 71}, 2777 (1993)]. When used recursively it improves simultaneously the node and phase. The algorithm is demonstrated to converge to the nearly exact solutions of model systems with periodic boundary conditions or applied magnetic fields. The method is also applied to obtain low energy excitations with magnetic field or periodic boundary conditions. The potential applications of this new method to study periodic, magnetic, and complex Hamiltonians are discussed.« less

Simulating polarized light scattering in terrestrial snow based on bicontinuous random medium and Monte Carlo ray tracing

NASA Astrophysics Data System (ADS)

Xiong, Chuan; Shi, Jiancheng

2014-01-01

To date, the light scattering models of snow consider very little about the real snow microstructures. The ideal spherical or other single shaped particle assumptions in previous snow light scattering models can cause error in light scattering modeling of snow and further cause errors in remote sensing inversion algorithms. This paper tries to build up a snow polarized reflectance model based on bicontinuous medium, with which the real snow microstructure is considered. The accurate specific surface area of bicontinuous medium can be analytically derived. The polarized Monte Carlo ray tracing technique is applied to the computer generated bicontinuous medium. With proper algorithms, the snow surface albedo, bidirectional reflectance distribution function (BRDF) and polarized BRDF can be simulated. The validation of model predicted spectral albedo and bidirectional reflectance factor (BRF) using experiment data shows good results. The relationship between snow surface albedo and snow specific surface area (SSA) were predicted, and this relationship can be used for future improvement of snow specific surface area (SSA) inversion algorithms. The model predicted polarized reflectance is validated and proved accurate, which can be further applied in polarized remote sensing.

Multiple Time-Step Dual-Hamiltonian Hybrid Molecular Dynamics — Monte Carlo Canonical Propagation Algorithm

PubMed Central

Weare, Jonathan; Dinner, Aaron R.; Roux, Benoît

2016-01-01

A multiple time-step integrator based on a dual Hamiltonian and a hybrid method combining molecular dynamics (MD) and Monte Carlo (MC) is proposed to sample systems in the canonical ensemble. The Dual Hamiltonian Multiple Time-Step (DHMTS) algorithm is based on two similar Hamiltonians: a computationally expensive one that serves as a reference and a computationally inexpensive one to which the workload is shifted. The central assumption is that the difference between the two Hamiltonians is slowly varying. Earlier work has shown that such dual Hamiltonian multiple time-step schemes effectively precondition nonlinear differential equations for dynamics by reformulating them into a recursive root finding problem that can be solved by propagating a correction term through an internal loop, analogous to RESPA. Of special interest in the present context, a hybrid MD-MC version of the DHMTS algorithm is introduced to enforce detailed balance via a Metropolis acceptance criterion and ensure consistency with the Boltzmann distribution. The Metropolis criterion suppresses the discretization errors normally associated with the propagation according to the computationally inexpensive Hamiltonian, treating the discretization error as an external work. Illustrative tests are carried out to demonstrate the effectiveness of the method. PMID:26918826

Stan : A Probabilistic Programming Language

DOE PAGES

Carpenter, Bob; Gelman, Andrew; Hoffman, Matthew D.; ...

2017-01-01

Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectationmore » propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can also be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.« less

SU-E-T-397: Evaluation of Planned Dose Distributions by Monte Carlo (0.5%) and Ray Tracing Algorithm for the Spinal Tumors with CyberKnife

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

Cho, H; Brindle, J; Hepel, J

2015-06-15

Purpose: To analyze and evaluate dose distribution between Ray Tracing (RT) and Monte Carlo (MC) algorithms of 0.5% uncertainty on a critical structure of spinal cord and gross target volume and planning target volume. Methods: Twenty four spinal tumor patients were treated with stereotactic body radiotherapy (SBRT) by CyberKnife in 2013 and 2014. The MC algorithm with 0.5% of uncertainty is used to recalculate the dose distribution for the treatment plan of the patients using the same beams, beam directions, and monitor units (MUs). Results: The prescription doses are uniformly larger for MC plans than RT except one case. Upmore » to a factor of 1.19 for 0.25cc threshold volume and 1.14 for 1.2cc threshold volume of dose differences are observed for the spinal cord. Conclusion: The MC recalculated dose distributions are larger than the original MC calculations for the spinal tumor cases. Based on the accuracy of the MC calculations, more radiation dose might be delivered to the tumor targets and spinal cords with the increase prescription dose.« less

Transforming high-dimensional potential energy surfaces into sum-of-products form using Monte Carlo methods

NASA Astrophysics Data System (ADS)

Schröder, Markus; Meyer, Hans-Dieter

2017-08-01

We propose a Monte Carlo method, "Monte Carlo Potfit," for transforming high-dimensional potential energy surfaces evaluated on discrete grid points into a sum-of-products form, more precisely into a Tucker form. To this end we use a variational ansatz in which we replace numerically exact integrals with Monte Carlo integrals. This largely reduces the numerical cost by avoiding the evaluation of the potential on all grid points and allows a treatment of surfaces up to 15-18 degrees of freedom. We furthermore show that the error made with this ansatz can be controlled and vanishes in certain limits. We present calculations on the potential of HFCO to demonstrate the features of the algorithm. To demonstrate the power of the method, we transformed a 15D potential of the protonated water dimer (Zundel cation) in a sum-of-products form and calculated the ground and lowest 26 vibrationally excited states of the Zundel cation with the multi-configuration time-dependent Hartree method.

Semi-stochastic full configuration interaction quantum Monte Carlo

NASA Astrophysics Data System (ADS)

Holmes, Adam; Petruzielo, Frank; Khadilkar, Mihir; Changlani, Hitesh; Nightingale, M. P.; Umrigar, C. J.

2012-02-01

In the recently proposed full configuration interaction quantum Monte Carlo (FCIQMC) [1,2], the ground state is projected out stochastically, using a population of walkers each of which represents a basis state in the Hilbert space spanned by Slater determinants. The infamous fermion sign problem manifests itself in the fact that walkers of either sign can be spawned on a given determinant. We propose an improvement on this method in the form of a hybrid stochastic/deterministic technique, which we expect will improve the efficiency of the algorithm by ameliorating the sign problem. We test the method on atoms and molecules, e.g., carbon, carbon dimer, N2 molecule, and stretched N2. [4pt] [1] Fermion Monte Carlo without fixed nodes: a Game of Life, death and annihilation in Slater Determinant space. George Booth, Alex Thom, Ali Alavi. J Chem Phys 131, 050106, (2009).[0pt] [2] Survival of the fittest: Accelerating convergence in full configuration-interaction quantum Monte Carlo. Deidre Cleland, George Booth, and Ali Alavi. J Chem Phys 132, 041103 (2010).

Stochastic, real-space, imaginary-time evaluation of third-order Feynman-Goldstone diagrams

NASA Astrophysics Data System (ADS)

Willow, Soohaeng Yoo; Hirata, So

2014-01-01

A new, alternative set of interpretation rules of Feynman-Goldstone diagrams for many-body perturbation theory is proposed, which translates diagrams into algebraic expressions suitable for direct Monte Carlo integrations. A vertex of a diagram is associated with a Coulomb interaction (rather than a two-electron integral) and an edge with the trace of a Green's function in real space and imaginary time. With these, 12 diagrams of third-order many-body perturbation (MP3) theory are converted into 20-dimensional integrals, which are then evaluated by a Monte Carlo method. It uses redundant walkers for convergence acceleration and a weight function for importance sampling in conjunction with the Metropolis algorithm. The resulting Monte Carlo MP3 method has low-rank polynomial size dependence of the operation cost, a negligible memory cost, and a naturally parallel computational kernel, while reproducing the correct correlation energies of small molecules within a few mEh after 106 Monte Carlo steps.

SU-E-T-405: Evaluation of the Raystation Electron Monte Carlo Algorithm for Varian Linear Accelerators

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

Sansourekidou, P; Allen, C

2015-06-15

Purpose: To evaluate the Raystation v4.51 Electron Monte Carlo algorithm for Varian Trilogy, IX and 2100 series linear accelerators and commission for clinical use. Methods: Seventy two water and forty air scans were acquired with a water tank in the form of profiles and depth doses, as requested by vendor. Data was imported into Rayphysics beam modeling module. Energy spectrum was modeled using seven parameters. Contamination photons were modeled using five parameters. Source phase space was modeled using six parameters. Calculations were performed in clinical version 4.51 and percent depth dose curves and profiles were extracted to be compared tomore » water tank measurements. Sensitivity tests were performed for all parameters. Grid size and particle histories were evaluated per energy for statistical uncertainty performance. Results: Model accuracy for air profiles is poor in the shoulder and penumbra region. However, model accuracy for water scans is acceptable. All energies and cones are within 2%/2mm for 90% of the points evaluated. Source phase space parameters have a cumulative effect. To achieve distributions with satisfactory smoothness level a 0.1cm grid and 3,000,000 particle histories were used for commissioning calculations. Calculation time was approximately 3 hours per energy. Conclusion: Raystation electron Monte Carlo is acceptable for clinical use for the Varian accelerators listed. Results are inferior to Elekta Electron Monte Carlo modeling. Known issues were reported to Raysearch and will be resolved in upcoming releases. Auto-modeling is limited to open cone depth dose curves and needs expansion.« less

Research on B Cell Algorithm for Learning to Rank Method Based on Parallel Strategy.

PubMed

Tian, Yuling; Zhang, Hongxian

2016-01-01

For the purposes of information retrieval, users must find highly relevant documents from within a system (and often a quite large one comprised of many individual documents) based on input query. Ranking the documents according to their relevance within the system to meet user needs is a challenging endeavor, and a hot research topic-there already exist several rank-learning methods based on machine learning techniques which can generate ranking functions automatically. This paper proposes a parallel B cell algorithm, RankBCA, for rank learning which utilizes a clonal selection mechanism based on biological immunity. The novel algorithm is compared with traditional rank-learning algorithms through experimentation and shown to outperform the others in respect to accuracy, learning time, and convergence rate; taken together, the experimental results show that the proposed algorithm indeed effectively and rapidly identifies optimal ranking functions.

Research on B Cell Algorithm for Learning to Rank Method Based on Parallel Strategy

PubMed Central

Tian, Yuling; Zhang, Hongxian

2016-01-01

For the purposes of information retrieval, users must find highly relevant documents from within a system (and often a quite large one comprised of many individual documents) based on input query. Ranking the documents according to their relevance within the system to meet user needs is a challenging endeavor, and a hot research topic–there already exist several rank-learning methods based on machine learning techniques which can generate ranking functions automatically. This paper proposes a parallel B cell algorithm, RankBCA, for rank learning which utilizes a clonal selection mechanism based on biological immunity. The novel algorithm is compared with traditional rank-learning algorithms through experimentation and shown to outperform the others in respect to accuracy, learning time, and convergence rate; taken together, the experimental results show that the proposed algorithm indeed effectively and rapidly identifies optimal ranking functions. PMID:27487242

Gradient descent learning algorithm overview: a general dynamical systems perspective.

PubMed

Baldi, P

1995-01-01

Gives a unified treatment of gradient descent learning algorithms for neural networks using a general framework of dynamical systems. This general approach organizes and simplifies all the known algorithms and results which have been originally derived for different problems (fixed point/trajectory learning), for different models (discrete/continuous), for different architectures (forward/recurrent), and using different techniques (backpropagation, variational calculus, adjoint methods, etc.). The general approach can also be applied to derive new algorithms. The author then briefly examines some of the complexity issues and limitations intrinsic to gradient descent learning. Throughout the paper, the author focuses on the problem of trajectory learning.

Minimizing the Discrepancy between Simulated and Historical Failures in Turbine Engines: A Simulation-Based Optimization Method (Postprint)

DTIC Science & Technology

2015-01-01

Procedure. The simulated annealing (SA) algorithm is a well-known local search metaheuristic used to address discrete, continuous, and multiobjective...design of experiments (DOE) to tune the parameters of the optimiza- tion algorithm . Section 5 shows the results of the case study. Finally, concluding... metaheuristic . The proposed method is broken down into two phases. Phase I consists of a Monte Carlo simulation to obtain the simulated percentage of failure

Light reflection models for computer graphics.

PubMed

Greenberg, D P

1989-04-14

During the past 20 years, computer graphic techniques for simulating the reflection of light have progressed so that today images of photorealistic quality can be produced. Early algorithms considered direct lighting only, but global illumination phenomena with indirect lighting, surface interreflections, and shadows can now be modeled with ray tracing, radiosity, and Monte Carlo simulations. This article describes the historical development of computer graphic algorithms for light reflection and pictorially illustrates what will be commonly available in the near future.

Slice sampling technique in Bayesian extreme of gold price modelling

NASA Astrophysics Data System (ADS)

Rostami, Mohammad; Adam, Mohd Bakri; Ibrahim, Noor Akma; Yahya, Mohamed Hisham

2013-09-01

In this paper, a simulation study of Bayesian extreme values by using Markov Chain Monte Carlo via slice sampling algorithm is implemented. We compared the accuracy of slice sampling with other methods for a Gumbel model. This study revealed that slice sampling algorithm offers more accurate and closer estimates with less RMSE than other methods . Finally we successfully employed this procedure to estimate the parameters of Malaysia extreme gold price from 2000 to 2011.

Network congestion control algorithm based on Actor-Critic reinforcement learning model

NASA Astrophysics Data System (ADS)

Xu, Tao; Gong, Lina; Zhang, Wei; Li, Xuhong; Wang, Xia; Pan, Wenwen

2018-04-01

Aiming at the network congestion control problem, a congestion control algorithm based on Actor-Critic reinforcement learning model is designed. Through the genetic algorithm in the congestion control strategy, the network congestion problems can be better found and prevented. According to Actor-Critic reinforcement learning, the simulation experiment of network congestion control algorithm is designed. The simulation experiments verify that the AQM controller can predict the dynamic characteristics of the network system. Moreover, the learning strategy is adopted to optimize the network performance, and the dropping probability of packets is adaptively adjusted so as to improve the network performance and avoid congestion. Based on the above finding, it is concluded that the network congestion control algorithm based on Actor-Critic reinforcement learning model can effectively avoid the occurrence of TCP network congestion.

Crystal-structure prediction via the Floppy-Box Monte Carlo algorithm: Method and application to hard (non)convex particles

NASA Astrophysics Data System (ADS)

de Graaf, Joost; Filion, Laura; Marechal, Matthieu; van Roij, René; Dijkstra, Marjolein

2012-12-01

In this paper, we describe the way to set up the floppy-box Monte Carlo (FBMC) method [L. Filion, M. Marechal, B. van Oorschot, D. Pelt, F. Smallenburg, and M. Dijkstra, Phys. Rev. Lett. 103, 188302 (2009), 10.1103/PhysRevLett.103.188302] to predict crystal-structure candidates for colloidal particles. The algorithm is explained in detail to ensure that it can be straightforwardly implemented on the basis of this text. The handling of hard-particle interactions in the FBMC algorithm is given special attention, as (soft) short-range and semi-long-range interactions can be treated in an analogous way. We also discuss two types of algorithms for checking for overlaps between polyhedra, the method of separating axes and a triangular-tessellation based technique. These can be combined with the FBMC method to enable crystal-structure prediction for systems composed of highly shape-anisotropic particles. Moreover, we present the results for the dense crystal structures predicted using the FBMC method for 159 (non)convex faceted particles, on which the findings in [J. de Graaf, R. van Roij, and M. Dijkstra, Phys. Rev. Lett. 107, 155501 (2011), 10.1103/PhysRevLett.107.155501] were based. Finally, we comment on the process of crystal-structure prediction itself and the choices that can be made in these simulations.

Modeling of light distribution in the brain for topographical imaging

NASA Astrophysics Data System (ADS)

Okada, Eiji; Hayashi, Toshiyuki; Kawaguchi, Hiroshi

2004-07-01

Multi-channel optical imaging system can obtain a topographical distribution of the activated region in the brain cortex by a simple mapping algorithm. Near-infrared light is strongly scattered in the head and the volume of tissue that contributes to the change in the optical signal detected with source-detector pair on the head surface is broadly distributed in the brain. This scattering effect results in poor resolution and contrast in the topographic image of the brain activity. We report theoretical investigations on the spatial resolution of the topographic imaging of the brain activity. The head model for the theoretical study consists of five layers that imitate the scalp, skull, subarachnoid space, gray matter and white matter. The light propagation in the head model is predicted by Monte Carlo simulation to obtain the spatial sensitivity profile for a source-detector pair. The source-detector pairs are one dimensionally arranged on the surface of the model and the distance between the adjoining source-detector pairs are varied from 4 mm to 32 mm. The change in detected intensity caused by the absorption change is obtained by Monte Carlo simulation. The position of absorption change is reconstructed by the conventional mapping algorithm and the reconstruction algorithm using the spatial sensitivity profiles. We discuss the effective interval between the source-detector pairs and the choice of reconstruction algorithms to improve the topographic images of brain activity.

Variance-reduced simulation of lattice discrete-time Markov chains with applications in reaction networks

NASA Astrophysics Data System (ADS)

Maginnis, P. A.; West, M.; Dullerud, G. E.

2016-10-01

We propose an algorithm to accelerate Monte Carlo simulation for a broad class of stochastic processes. Specifically, the class of countable-state, discrete-time Markov chains driven by additive Poisson noise, or lattice discrete-time Markov chains. In particular, this class includes simulation of reaction networks via the tau-leaping algorithm. To produce the speedup, we simulate pairs of fair-draw trajectories that are negatively correlated. Thus, when averaged, these paths produce an unbiased Monte Carlo estimator that has reduced variance and, therefore, reduced error. Numerical results for three example systems included in this work demonstrate two to four orders of magnitude reduction of mean-square error. The numerical examples were chosen to illustrate different application areas and levels of system complexity. The areas are: gene expression (affine state-dependent rates), aerosol particle coagulation with emission and human immunodeficiency virus infection (both with nonlinear state-dependent rates). Our algorithm views the system dynamics as a ;black-box;, i.e., we only require control of pseudorandom number generator inputs. As a result, typical codes can be retrofitted with our algorithm using only minor changes. We prove several analytical results. Among these, we characterize the relationship of covariances between paths in the general nonlinear state-dependent intensity rates case, and we prove variance reduction of mean estimators in the special case of affine intensity rates.

On an adaptive preconditioned Crank-Nicolson MCMC algorithm for infinite dimensional Bayesian inference

NASA Astrophysics Data System (ADS)

Hu, Zixi; Yao, Zhewei; Li, Jinglai

2017-03-01

Many scientific and engineering problems require to perform Bayesian inference for unknowns of infinite dimension. In such problems, many standard Markov Chain Monte Carlo (MCMC) algorithms become arbitrary slow under the mesh refinement, which is referred to as being dimension dependent. To this end, a family of dimensional independent MCMC algorithms, known as the preconditioned Crank-Nicolson (pCN) methods, were proposed to sample the infinite dimensional parameters. In this work we develop an adaptive version of the pCN algorithm, where the covariance operator of the proposal distribution is adjusted based on sampling history to improve the simulation efficiency. We show that the proposed algorithm satisfies an important ergodicity condition under some mild assumptions. Finally we provide numerical examples to demonstrate the performance of the proposed method.

Algorithm Visualization System for Teaching Spatial Data Algorithms

ERIC Educational Resources Information Center

Nikander, Jussi; Helminen, Juha; Korhonen, Ari

2010-01-01

TRAKLA2 is a web-based learning environment for data structures and algorithms. The system delivers automatically assessed algorithm simulation exercises that are solved using a graphical user interface. In this work, we introduce a novel learning environment for spatial data algorithms, SDA-TRAKLA2, which has been implemented on top of the…

Constrained Metric Learning by Permutation Inducing Isometries.

PubMed

Bosveld, Joel; Mahmood, Arif; Huynh, Du Q; Noakes, Lyle

2016-01-01

The choice of metric critically affects the performance of classification and clustering algorithms. Metric learning algorithms attempt to improve performance, by learning a more appropriate metric. Unfortunately, most of the current algorithms learn a distance function which is not invariant to rigid transformations of images. Therefore, the distances between two images and their rigidly transformed pair may differ, leading to inconsistent classification or clustering results. We propose to constrain the learned metric to be invariant to the geometry preserving transformations of images that induce permutations in the feature space. The constraint that these transformations are isometries of the metric ensures consistent results and improves accuracy. Our second contribution is a dimension reduction technique that is consistent with the isometry constraints. Our third contribution is the formulation of the isometry constrained logistic discriminant metric learning (IC-LDML) algorithm, by incorporating the isometry constraints within the objective function of the LDML algorithm. The proposed algorithm is compared with the existing techniques on the publicly available labeled faces in the wild, viewpoint-invariant pedestrian recognition, and Toy Cars data sets. The IC-LDML algorithm has outperformed existing techniques for the tasks of face recognition, person identification, and object classification by a significant margin.

Time-optimal trajectory planning for underactuated spacecraft using a hybrid particle swarm optimization algorithm

NASA Astrophysics Data System (ADS)

Zhuang, Yufei; Huang, Haibin

2014-02-01

A hybrid algorithm combining particle swarm optimization (PSO) algorithm with the Legendre pseudospectral method (LPM) is proposed for solving time-optimal trajectory planning problem of underactuated spacecrafts. At the beginning phase of the searching process, an initialization generator is constructed by the PSO algorithm due to its strong global searching ability and robustness to random initial values, however, PSO algorithm has a disadvantage that its convergence rate around the global optimum is slow. Then, when the change in fitness function is smaller than a predefined value, the searching algorithm is switched to the LPM to accelerate the searching process. Thus, with the obtained solutions by the PSO algorithm as a set of proper initial guesses, the hybrid algorithm can find a global optimum more quickly and accurately. 200 Monte Carlo simulations results demonstrate that the proposed hybrid PSO-LPM algorithm has greater advantages in terms of global searching capability and convergence rate than both single PSO algorithm and LPM algorithm. Moreover, the PSO-LPM algorithm is also robust to random initial values.

An incremental approach to genetic-algorithms-based classification.

PubMed

Guan, Sheng-Uei; Zhu, Fangming

2005-04-01

Incremental learning has been widely addressed in the machine learning literature to cope with learning tasks where the learning environment is ever changing or training samples become available over time. However, most research work explores incremental learning with statistical algorithms or neural networks, rather than evolutionary algorithms. The work in this paper employs genetic algorithms (GAs) as basic learning algorithms for incremental learning within one or more classifier agents in a multiagent environment. Four new approaches with different initialization schemes are proposed. They keep the old solutions and use an "integration" operation to integrate them with new elements to accommodate new attributes, while biased mutation and crossover operations are adopted to further evolve a reinforced solution. The simulation results on benchmark classification data sets show that the proposed approaches can deal with the arrival of new input attributes and integrate them with the original input space. It is also shown that the proposed approaches can be successfully used for incremental learning and improve classification rates as compared to the retraining GA. Possible applications for continuous incremental training and feature selection are also discussed.

Data-driven advice for applying machine learning to bioinformatics problems

PubMed Central

Olson, Randal S.; La Cava, William; Mustahsan, Zairah; Varik, Akshay; Moore, Jason H.

2017-01-01

As the bioinformatics field grows, it must keep pace not only with new data but with new algorithms. Here we contribute a thorough analysis of 13 state-of-the-art, commonly used machine learning algorithms on a set of 165 publicly available classification problems in order to provide data-driven algorithm recommendations to current researchers. We present a number of statistical and visual comparisons of algorithm performance and quantify the effect of model selection and algorithm tuning for each algorithm and dataset. The analysis culminates in the recommendation of five algorithms with hyperparameters that maximize classifier performance across the tested problems, as well as general guidelines for applying machine learning to supervised classification problems. PMID:29218881

Designing the Architecture of Hierachical Neural Networks Model Attention, Learning and Goal-Oriented Behavior

DTIC Science & Technology

1993-12-31

19,23,25,26,27,28,32,33,35,41]) - A new cost function is postulated and an algorithm that employs this cost function is proposed for the learning of...updates the controller parameters from time to time [53]. The learning control algorithm consist of updating the parameter estimates as used in the...proposed cost function with the other learning type algorithms , such as based upon learning of iterative tasks [Kawamura-85], variable structure

EFFICIENT MODEL-FITTING AND MODEL-COMPARISON FOR HIGH-DIMENSIONAL BAYESIAN GEOSTATISTICAL MODELS. (R826887)

EPA Science Inventory

Geostatistical models are appropriate for spatially distributed data measured at irregularly spaced locations. We propose an efficient Markov chain Monte Carlo (MCMC) algorithm for fitting Bayesian geostatistical models with substantial numbers of unknown parameters to sizable...

Behavioral Profiling of Scada Network Traffic Using Machine Learning Algorithms

DTIC Science & Technology

2014-03-27

BEHAVIORAL PROFILING OF SCADA NETWORK TRAFFIC USING MACHINE LEARNING ALGORITHMS THESIS Jessica R. Werling, Captain, USAF AFIT-ENG-14-M-81 DEPARTMENT...subject to copyright protection in the United States. AFIT-ENG-14-M-81 BEHAVIORAL PROFILING OF SCADA NETWORK TRAFFIC USING MACHINE LEARNING ...AFIT-ENG-14-M-81 BEHAVIORAL PROFILING OF SCADA NETWORK TRAFFIC USING MACHINE LEARNING ALGORITHMS Jessica R. Werling, B.S.C.S. Captain, USAF Approved

Super-resolution reconstruction of MR image with a novel residual learning network algorithm

NASA Astrophysics Data System (ADS)

Shi, Jun; Liu, Qingping; Wang, Chaofeng; Zhang, Qi; Ying, Shihui; Xu, Haoyu

2018-04-01

Spatial resolution is one of the key parameters of magnetic resonance imaging (MRI). The image super-resolution (SR) technique offers an alternative approach to improve the spatial resolution of MRI due to its simplicity. Convolutional neural networks (CNN)-based SR algorithms have achieved state-of-the-art performance, in which the global residual learning (GRL) strategy is now commonly used due to its effectiveness for learning image details for SR. However, the partial loss of image details usually happens in a very deep network due to the degradation problem. In this work, we propose a novel residual learning-based SR algorithm for MRI, which combines both multi-scale GRL and shallow network block-based local residual learning (LRL). The proposed LRL module works effectively in capturing high-frequency details by learning local residuals. One simulated MRI dataset and two real MRI datasets have been used to evaluate our algorithm. The experimental results show that the proposed SR algorithm achieves superior performance to all of the other compared CNN-based SR algorithms in this work.

Hardware Acceleration of Adaptive Neural Algorithms.

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

James, Conrad D.

As tradit ional numerical computing has faced challenges, researchers have turned towards alternative computing approaches to reduce power - per - computation metrics and improve algorithm performance. Here, we describe an approach towards non - conventional computing that strengthens the connection between machine learning and neuroscience concepts. The Hardware Acceleration of Adaptive Neural Algorithms (HAANA) project ha s develop ed neural machine learning algorithms and hardware for applications in image processing and cybersecurity. While machine learning methods are effective at extracting relevant features from many types of data, the effectiveness of these algorithms degrades when subjected to real - worldmore » conditions. Our team has generated novel neural - inspired approa ches to improve the resiliency and adaptability of machine learning algorithms. In addition, we have also designed and fabricated hardware architectures and microelectronic devices specifically tuned towards the training and inference operations of neural - inspired algorithms. Finally, our multi - scale simulation framework allows us to assess the impact of microelectronic device properties on algorithm performance.« less

A new concept of pencil beam dose calculation for 40-200 keV photons using analytical dose kernels.

PubMed

Bartzsch, Stefan; Oelfke, Uwe

2013-11-01

The advent of widespread kV-cone beam computer tomography in image guided radiation therapy and special therapeutic application of keV photons, e.g., in microbeam radiation therapy (MRT) require accurate and fast dose calculations for photon beams with energies between 40 and 200 keV. Multiple photon scattering originating from Compton scattering and the strong dependence of the photoelectric cross section on the atomic number of the interacting tissue render these dose calculations by far more challenging than the ones established for corresponding MeV beams. That is why so far developed analytical models of kV photon dose calculations fail to provide the required accuracy and one has to rely on time consuming Monte Carlo simulation techniques. In this paper, the authors introduce a novel analytical approach for kV photon dose calculations with an accuracy that is almost comparable to the one of Monte Carlo simulations. First, analytical point dose and pencil beam kernels are derived for homogeneous media and compared to Monte Carlo simulations performed with the Geant4 toolkit. The dose contributions are systematically separated into contributions from the relevant orders of multiple photon scattering. Moreover, approximate scaling laws for the extension of the algorithm to inhomogeneous media are derived. The comparison of the analytically derived dose kernels in water showed an excellent agreement with the Monte Carlo method. Calculated values deviate less than 5% from Monte Carlo derived dose values, for doses above 1% of the maximum dose. The analytical structure of the kernels allows adaption to arbitrary materials and photon spectra in the given energy range of 40-200 keV. The presented analytical methods can be employed in a fast treatment planning system for MRT. In convolution based algorithms dose calculation times can be reduced to a few minutes.

A Monte Carlo method for the simulation of coagulation and nucleation based on weighted particles and the concepts of stochastic resolution and merging

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

Kotalczyk, G., E-mail: Gregor.Kotalczyk@uni-due.de; Kruis, F.E.

Monte Carlo simulations based on weighted simulation particles can solve a variety of population balance problems and allow thus to formulate a solution-framework for many chemical engineering processes. This study presents a novel concept for the calculation of coagulation rates of weighted Monte Carlo particles by introducing a family of transformations to non-weighted Monte Carlo particles. The tuning of the accuracy (named ‘stochastic resolution’ in this paper) of those transformations allows the construction of a constant-number coagulation scheme. Furthermore, a parallel algorithm for the inclusion of newly formed Monte Carlo particles due to nucleation is presented in the scope ofmore » a constant-number scheme: the low-weight merging. This technique is found to create significantly less statistical simulation noise than the conventional technique (named ‘random removal’ in this paper). Both concepts are combined into a single GPU-based simulation method which is validated by comparison with the discrete-sectional simulation technique. Two test models describing a constant-rate nucleation coupled to a simultaneous coagulation in 1) the free-molecular regime or 2) the continuum regime are simulated for this purpose.« less

Monte Carlo: in the beginning and some great expectations

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

Metropolis, N.

1985-01-01

The central theme will be on the historical setting and origins of the Monte Carlo Method. The scene was post-war Los Alamos Scientific Laboratory. There was an inevitability about the Monte Carlo Event: the ENIAC had recently enjoyed its meteoric rise (on a classified Los Alamos problem); Stan Ulam had returned to Los Alamos; John von Neumann was a frequent visitor. Techniques, algorithms, and applications developed rapidly at Los Alamos. Soon, the fascination of the Method reached wider horizons. The first paper was submitted for publication in the spring of 1949. In the summer of 1949, the first open conferencemore » was held at the University of California at Los Angeles. Of some interst perhaps is an account of Fermi's earlier, independent application in neutron moderation studies while at the University of Rome. The quantum leap expected with the advent of massively parallel processors will provide stimuli for very ambitious applications of the Monte Carlo Method in disciplines ranging from field theories to cosmology, including more realistic models in the neurosciences. A structure of multi-instruction sets for parallel processing is ideally suited for the Monte Carlo approach. One may even hope for a modest hardening of the soft sciences.« less

On the utility of graphics cards to perform massively parallel simulation of advanced Monte Carlo methods

PubMed Central

Lee, Anthony; Yau, Christopher; Giles, Michael B.; Doucet, Arnaud; Holmes, Christopher C.

2011-01-01

We present a case-study on the utility of graphics cards to perform massively parallel simulation of advanced Monte Carlo methods. Graphics cards, containing multiple Graphics Processing Units (GPUs), are self-contained parallel computational devices that can be housed in conventional desktop and laptop computers and can be thought of as prototypes of the next generation of many-core processors. For certain classes of population-based Monte Carlo algorithms they offer massively parallel simulation, with the added advantage over conventional distributed multi-core processors that they are cheap, easily accessible, easy to maintain, easy to code, dedicated local devices with low power consumption. On a canonical set of stochastic simulation examples including population-based Markov chain Monte Carlo methods and Sequential Monte Carlo methods, we nd speedups from 35 to 500 fold over conventional single-threaded computer code. Our findings suggest that GPUs have the potential to facilitate the growth of statistical modelling into complex data rich domains through the availability of cheap and accessible many-core computation. We believe the speedup we observe should motivate wider use of parallelizable simulation methods and greater methodological attention to their design. PMID:22003276

Using human brain activity to guide machine learning.

PubMed

Fong, Ruth C; Scheirer, Walter J; Cox, David D

2018-03-29

Machine learning is a field of computer science that builds algorithms that learn. In many cases, machine learning algorithms are used to recreate a human ability like adding a caption to a photo, driving a car, or playing a game. While the human brain has long served as a source of inspiration for machine learning, little effort has been made to directly use data collected from working brains as a guide for machine learning algorithms. Here we demonstrate a new paradigm of "neurally-weighted" machine learning, which takes fMRI measurements of human brain activity from subjects viewing images, and infuses these data into the training process of an object recognition learning algorithm to make it more consistent with the human brain. After training, these neurally-weighted classifiers are able to classify images without requiring any additional neural data. We show that our neural-weighting approach can lead to large performance gains when used with traditional machine vision features, as well as to significant improvements with already high-performing convolutional neural network features. The effectiveness of this approach points to a path forward for a new class of hybrid machine learning algorithms which take both inspiration and direct constraints from neuronal data.

Javascript Library for Developing Interactive Micro-Level Animations for Teaching and Learning Algorithms on One-Dimensional Arrays

ERIC Educational Resources Information Center

Végh, Ladislav

2016-01-01

The first data structure that first-year undergraduate students learn during the programming and algorithms courses is the one-dimensional array. For novice programmers, it might be hard to understand different algorithms on arrays (e.g. searching, mirroring, sorting algorithms), because the algorithms dynamically change the values of elements. In…

Inductive learning of thyroid functional states using the ID3 algorithm. The effect of poor examples on the learning result.

PubMed

Forsström, J

1992-01-01

The ID3 algorithm for inductive learning was tested using preclassified material for patients suspected to have a thyroid illness. Classification followed a rule-based expert system for the diagnosis of thyroid function. Thus, the knowledge to be learned was limited to the rules existing in the knowledge base of that expert system. The learning capability of the ID3 algorithm was tested with an unselected learning material (with some inherent missing data) and with a selected learning material (no missing data). The selected learning material was a subgroup which formed a part of the unselected learning material. When the number of learning cases was increased, the accuracy of the program improved. When the learning material was large enough, an increase in the learning material did not improve the results further. A better learning result was achieved with the selected learning material not including missing data as compared to unselected learning material. With this material we demonstrate a weakness in the ID3 algorithm: it can not find available information from good example cases if we add poor examples to the data.

Privacy-preserving backpropagation neural network learning.

PubMed

Chen, Tingting; Zhong, Sheng

2009-10-01

With the development of distributed computing environment , many learning problems now have to deal with distributed input data. To enhance cooperations in learning, it is important to address the privacy concern of each data holder by extending the privacy preservation notion to original learning algorithms. In this paper, we focus on preserving the privacy in an important learning model, multilayer neural networks. We present a privacy-preserving two-party distributed algorithm of backpropagation which allows a neural network to be trained without requiring either party to reveal her data to the other. We provide complete correctness and security analysis of our algorithms. The effectiveness of our algorithms is verified by experiments on various real world data sets.

Learning Cue Phrase Patterns from Radiology Reports Using a Genetic Algorithm

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

Patton, Robert M; Beckerman, Barbara G; Potok, Thomas E

2009-01-01

Various computer-assisted technologies have been developed to assist radiologists in detecting cancer; however, the algorithms still lack high degrees of sensitivity and specificity, and must undergo machine learning against a training set with known pathologies in order to further refine the algorithms with higher validity of truth. This work describes an approach to learning cue phrase patterns in radiology reports that utilizes a genetic algorithm (GA) as the learning method. The approach described here successfully learned cue phrase patterns for two distinct classes of radiology reports. These patterns can then be used as a basis for automatically categorizing, clustering, ormore » retrieving relevant data for the user.« less

New Tools to Prepare ACE Cross-section Files for MCNP Analytic Test Problems

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

Brown, Forrest B.

Monte Carlo calculations using one-group cross sections, multigroup cross sections, or simple continuous energy cross sections are often used to: (1) verify production codes against known analytical solutions, (2) verify new methods and algorithms that do not involve detailed collision physics, (3) compare Monte Carlo calculation methods with deterministic methods, and (4) teach fundamentals to students. In this work we describe 2 new tools for preparing the ACE cross-section files to be used by MCNP ® for these analytic test problems, simple_ace.pl and simple_ace_mg.pl.

Monte Carlo errors with less errors

NASA Astrophysics Data System (ADS)

Wolff, Ulli; Alpha Collaboration

2004-01-01

We explain in detail how to estimate mean values and assess statistical errors for arbitrary functions of elementary observables in Monte Carlo simulations. The method is to estimate and sum the relevant autocorrelation functions, which is argued to produce more certain error estimates than binning techniques and hence to help toward a better exploitation of expensive simulations. An effective integrated autocorrelation time is computed which is suitable to benchmark efficiencies of simulation algorithms with regard to specific observables of interest. A Matlab code is offered for download that implements the method. It can also combine independent runs (replica) allowing to judge their consistency.

Markov Chain Monte Carlo Used in Parameter Inference of Magnetic Resonance Spectra

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

Hock, Kiel; Earle, Keith

2016-02-06

In this paper, we use Boltzmann statistics and the maximum likelihood distribution derived from Bayes’ Theorem to infer parameter values for a Pake Doublet Spectrum, a lineshape of historical significance and contemporary relevance for determining distances between interacting magnetic dipoles. A Metropolis Hastings Markov Chain Monte Carlo algorithm is implemented and designed to find the optimum parameter set and to estimate parameter uncertainties. In conclusion, the posterior distribution allows us to define a metric on parameter space that induces a geometry with negative curvature that affects the parameter uncertainty estimates, particularly for spectra with low signal to noise.

Driven-dissipative quantum Monte Carlo method for open quantum systems

NASA Astrophysics Data System (ADS)

Nagy, Alexandra; Savona, Vincenzo

2018-05-01

We develop a real-time full configuration-interaction quantum Monte Carlo approach to model driven-dissipative open quantum systems with Markovian system-bath coupling. The method enables stochastic sampling of the Liouville-von Neumann time evolution of the density matrix thanks to a massively parallel algorithm, thus providing estimates of observables on the nonequilibrium steady state. We present the underlying theory and introduce an initiator technique and importance sampling to reduce the statistical error. Finally, we demonstrate the efficiency of our approach by applying it to the driven-dissipative two-dimensional X Y Z spin-1/2 model on a lattice.

Computational characterization of HPGe detectors usable for a wide variety of source geometries by using Monte Carlo simulation and a multi-objective evolutionary algorithm

NASA Astrophysics Data System (ADS)

Guerra, J. G.; Rubiano, J. G.; Winter, G.; Guerra, A. G.; Alonso, H.; Arnedo, M. A.; Tejera, A.; Martel, P.; Bolivar, J. P.

2017-06-01

In this work, we have developed a computational methodology for characterizing HPGe detectors by implementing in parallel a multi-objective evolutionary algorithm, together with a Monte Carlo simulation code. The evolutionary algorithm is used for searching the geometrical parameters of a model of detector by minimizing the differences between the efficiencies calculated by Monte Carlo simulation and two reference sets of Full Energy Peak Efficiencies (FEPEs) corresponding to two given sample geometries, a beaker of small diameter laid over the detector window and a beaker of large capacity which wrap the detector. This methodology is a generalization of a previously published work, which was limited to beakers placed over the window of the detector with a diameter equal or smaller than the crystal diameter, so that the crystal mount cap (which surround the lateral surface of the crystal), was not considered in the detector model. The generalization has been accomplished not only by including such a mount cap in the model, but also using multi-objective optimization instead of mono-objective, with the aim of building a model sufficiently accurate for a wider variety of beakers commonly used for the measurement of environmental samples by gamma spectrometry, like for instance, Marinellis, Petris, or any other beaker with a diameter larger than the crystal diameter, for which part of the detected radiation have to pass through the mount cap. The proposed methodology has been applied to an HPGe XtRa detector, providing a model of detector which has been successfully verificated for different source-detector geometries and materials and experimentally validated using CRMs.

Optimizing the learning rate for adaptive estimation of neural encoding models

PubMed Central

2018-01-01

Closed-loop neurotechnologies often need to adaptively learn an encoding model that relates the neural activity to the brain state, and is used for brain state decoding. The speed and accuracy of adaptive learning algorithms are critically affected by the learning rate, which dictates how fast model parameters are updated based on new observations. Despite the importance of the learning rate, currently an analytical approach for its selection is largely lacking and existing signal processing methods vastly tune it empirically or heuristically. Here, we develop a novel analytical calibration algorithm for optimal selection of the learning rate in adaptive Bayesian filters. We formulate the problem through a fundamental trade-off that learning rate introduces between the steady-state error and the convergence time of the estimated model parameters. We derive explicit functions that predict the effect of learning rate on error and convergence time. Using these functions, our calibration algorithm can keep the steady-state parameter error covariance smaller than a desired upper-bound while minimizing the convergence time, or keep the convergence time faster than a desired value while minimizing the error. We derive the algorithm both for discrete-valued spikes modeled as point processes nonlinearly dependent on the brain state, and for continuous-valued neural recordings modeled as Gaussian processes linearly dependent on the brain state. Using extensive closed-loop simulations, we show that the analytical solution of the calibration algorithm accurately predicts the effect of learning rate on parameter error and convergence time. Moreover, the calibration algorithm allows for fast and accurate learning of the encoding model and for fast convergence of decoding to accurate performance. Finally, larger learning rates result in inaccurate encoding models and decoders, and smaller learning rates delay their convergence. The calibration algorithm provides a novel analytical approach to predictably achieve a desired level of error and convergence time in adaptive learning, with application to closed-loop neurotechnologies and other signal processing domains. PMID:29813069

Optimizing the learning rate for adaptive estimation of neural encoding models.

PubMed

Hsieh, Han-Lin; Shanechi, Maryam M

2018-05-01

Closed-loop neurotechnologies often need to adaptively learn an encoding model that relates the neural activity to the brain state, and is used for brain state decoding. The speed and accuracy of adaptive learning algorithms are critically affected by the learning rate, which dictates how fast model parameters are updated based on new observations. Despite the importance of the learning rate, currently an analytical approach for its selection is largely lacking and existing signal processing methods vastly tune it empirically or heuristically. Here, we develop a novel analytical calibration algorithm for optimal selection of the learning rate in adaptive Bayesian filters. We formulate the problem through a fundamental trade-off that learning rate introduces between the steady-state error and the convergence time of the estimated model parameters. We derive explicit functions that predict the effect of learning rate on error and convergence time. Using these functions, our calibration algorithm can keep the steady-state parameter error covariance smaller than a desired upper-bound while minimizing the convergence time, or keep the convergence time faster than a desired value while minimizing the error. We derive the algorithm both for discrete-valued spikes modeled as point processes nonlinearly dependent on the brain state, and for continuous-valued neural recordings modeled as Gaussian processes linearly dependent on the brain state. Using extensive closed-loop simulations, we show that the analytical solution of the calibration algorithm accurately predicts the effect of learning rate on parameter error and convergence time. Moreover, the calibration algorithm allows for fast and accurate learning of the encoding model and for fast convergence of decoding to accurate performance. Finally, larger learning rates result in inaccurate encoding models and decoders, and smaller learning rates delay their convergence. The calibration algorithm provides a novel analytical approach to predictably achieve a desired level of error and convergence time in adaptive learning, with application to closed-loop neurotechnologies and other signal processing domains.

Exploration of the association rules mining technique for the signal detection of adverse drug events in spontaneous reporting systems.

PubMed

Wang, Chao; Guo, Xiao-Jing; Xu, Jin-Fang; Wu, Cheng; Sun, Ya-Lin; Ye, Xiao-Fei; Qian, Wei; Ma, Xiu-Qiang; Du, Wen-Min; He, Jia

2012-01-01

The detection of signals of adverse drug events (ADEs) has increased because of the use of data mining algorithms in spontaneous reporting systems (SRSs). However, different data mining algorithms have different traits and conditions for application. The objective of our study was to explore the application of association rule (AR) mining in ADE signal detection and to compare its performance with that of other algorithms. Monte Carlo simulation was applied to generate drug-ADE reports randomly according to the characteristics of SRS datasets. Thousand simulated datasets were mined by AR and other algorithms. On average, 108,337 reports were generated by the Monte Carlo simulation. Based on the predefined criterion that 10% of the drug-ADE combinations were true signals, with RR equaling to 10, 4.9, 1.5, and 1.2, AR detected, on average, 284 suspected associations with a minimum support of 3 and a minimum lift of 1.2. The area under the receiver operating characteristic (ROC) curve of the AR was 0.788, which was equivalent to that shown for other algorithms. Additionally, AR was applied to reports submitted to the Shanghai SRS in 2009. Five hundred seventy combinations were detected using AR from 24,297 SRS reports, and they were compared with recognized ADEs identified by clinical experts and various other sources. AR appears to be an effective method for ADE signal detection, both in simulated and real SRS datasets. The limitations of this method exposed in our study, i.e., a non-uniform thresholds setting and redundant rules, require further research.

An Atmospheric Guidance Algorithm Testbed for the Mars Surveyor Program 2001 Orbiter and Lander

NASA Technical Reports Server (NTRS)

Striepe, Scott A.; Queen, Eric M.; Powell, Richard W.; Braun, Robert D.; Cheatwood, F. McNeil; Aguirre, John T.; Sachi, Laura A.; Lyons, Daniel T.

1998-01-01

An Atmospheric Flight Team was formed by the Mars Surveyor Program '01 mission office to develop aerocapture and precision landing testbed simulations and candidate guidance algorithms. Three- and six-degree-of-freedom Mars atmospheric flight simulations have been developed for testing, evaluation, and analysis of candidate guidance algorithms for the Mars Surveyor Program 2001 Orbiter and Lander. These simulations are built around the Program to Optimize Simulated Trajectories. Subroutines were supplied by Atmospheric Flight Team members for modeling the Mars atmosphere, spacecraft control system, aeroshell aerodynamic characteristics, and other Mars 2001 mission specific models. This paper describes these models and their perturbations applied during Monte Carlo analyses to develop, test, and characterize candidate guidance algorithms.

LLNL Mercury Project Trinity Open Science Final Report

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

Brantley, Patrick; Dawson, Shawn; McKinley, Scott

2016-04-20

The Mercury Monte Carlo particle transport code developed at Lawrence Livermore National Laboratory (LLNL) is used to simulate the transport of radiation through urban environments. These challenging calculations include complicated geometries and require significant computational resources to complete. As a result, a question arises as to the level of convergence of the calculations with Monte Carlo simulation particle count. In the Trinity Open Science calculations, one main focus was to investigate convergence of the relevant simulation quantities with Monte Carlo particle count to assess the current simulation methodology. Both for this application space but also of more general applicability, wemore » also investigated the impact of code algorithms on parallel scaling on the Trinity machine as well as the utilization of the Trinity DataWarp burst buffer technology in Mercury via the LLNL Scalable Checkpoint/Restart (SCR) library.« less

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

Costa, Liborio I., E-mail: liborio78@gmail.com

A new Markov Chain Monte Carlo method for simulating the dynamics of particle systems characterized by hard-core interactions is introduced. In contrast to traditional Kinetic Monte Carlo approaches, where the state of the system is associated with minima in the energy landscape, in the proposed method, the state of the system is associated with the set of paths traveled by the atoms and the transition probabilities for an atom to be displaced are proportional to the corresponding velocities. In this way, the number of possible state-to-state transitions is reduced to a discrete set, and a direct link between the Montemore » Carlo time step and true physical time is naturally established. The resulting rejection-free algorithm is validated against event-driven molecular dynamics: the equilibrium and non-equilibrium dynamics of hard disks converge to the exact results with decreasing displacement size.« less

A calibrated Monte Carlo approach to quantify the impacts of misorientation and different driving forces on texture development

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

Liangzhe Zhang; Anthony D. Rollett; Timothy Bartel

2012-02-01

A calibrated Monte Carlo (cMC) approach, which quantifies grain boundary kinetics within a generic setting, is presented. The influence of misorientation is captured by adding a scaling coefficient in the spin flipping probability equation, while the contribution of different driving forces is weighted using a partition function. The calibration process relies on the established parametric links between Monte Carlo (MC) and sharp-interface models. The cMC algorithm quantifies microstructural evolution under complex thermomechanical environments and remedies some of the difficulties associated with conventional MC models. After validation, the cMC approach is applied to quantify the texture development of polycrystalline materials withmore » influences of misorientation and inhomogeneous bulk energy across grain boundaries. The results are in good agreement with theory and experiments.« less

Advancing X-ray scattering metrology using inverse genetic algorithms.

PubMed

Hannon, Adam F; Sunday, Daniel F; Windover, Donald; Kline, R Joseph

2016-01-01

We compare the speed and effectiveness of two genetic optimization algorithms to the results of statistical sampling via a Markov chain Monte Carlo algorithm to find which is the most robust method for determining real space structure in periodic gratings measured using critical dimension small angle X-ray scattering. Both a covariance matrix adaptation evolutionary strategy and differential evolution algorithm are implemented and compared using various objective functions. The algorithms and objective functions are used to minimize differences between diffraction simulations and measured diffraction data. These simulations are parameterized with an electron density model known to roughly correspond to the real space structure of our nanogratings. The study shows that for X-ray scattering data, the covariance matrix adaptation coupled with a mean-absolute error log objective function is the most efficient combination of algorithm and goodness of fit criterion for finding structures with little foreknowledge about the underlying fine scale structure features of the nanograting.

Trans-algorithmic nature of learning in biological systems.

PubMed

Shimansky, Yury P

2018-05-02

Learning ability is a vitally important, distinctive property of biological systems, which provides dynamic stability in non-stationary environments. Although several different types of learning have been successfully modeled using a universal computer, in general, learning cannot be described by an algorithm. In other words, algorithmic approach to describing the functioning of biological systems is not sufficient for adequate grasping of what is life. Since biosystems are parts of the physical world, one might hope that adding some physical mechanisms and principles to the concept of algorithm could provide extra possibilities for describing learning in its full generality. However, a straightforward approach to that through the so-called physical hypercomputation so far has not been successful. Here an alternative approach is proposed. Biosystems are described as achieving enumeration of possible physical compositions though random incremental modifications inflicted on them by active operating resources (AORs) in the environment. Biosystems learn through algorithmic regulation of the intensity of the above modifications according to a specific optimality criterion. From the perspective of external observers, biosystems move in the space of different algorithms driven by random modifications imposed by the environmental AORs. A particular algorithm is only a snapshot of that motion, while the motion itself is essentially trans-algorithmic. In this conceptual framework, death of unfit members of a population, for example, is viewed as a trans-algorithmic modification made in the population as a biosystem by environmental AORs. Numerous examples of AOR utilization in biosystems of different complexity, from viruses to multicellular organisms, are provided.

A study of metaheuristic algorithms for high dimensional feature selection on microarray data

NASA Astrophysics Data System (ADS)

Dankolo, Muhammad Nasiru; Radzi, Nor Haizan Mohamed; Sallehuddin, Roselina; Mustaffa, Noorfa Haszlinna

2017-11-01

Microarray systems enable experts to examine gene profile at molecular level using machine learning algorithms. It increases the potentials of classification and diagnosis of many diseases at gene expression level. Though, numerous difficulties may affect the efficiency of machine learning algorithms which includes vast number of genes features comprised in the original data. Many of these features may be unrelated to the intended analysis. Therefore, feature selection is necessary to be performed in the data pre-processing. Many feature selection algorithms are developed and applied on microarray which including the metaheuristic optimization algorithms. This paper discusses the application of the metaheuristics algorithms for feature selection in microarray dataset. This study reveals that, the algorithms have yield an interesting result with limited resources thereby saving computational expenses of machine learning algorithms.

Successful attack on permutation-parity-machine-based neural cryptography.

PubMed

Seoane, Luís F; Ruttor, Andreas

2012-02-01

An algorithm is presented which implements a probabilistic attack on the key-exchange protocol based on permutation parity machines. Instead of imitating the synchronization of the communicating partners, the strategy consists of a Monte Carlo method to sample the space of possible weights during inner rounds and an analytic approach to convey the extracted information from one outer round to the next one. The results show that the protocol under attack fails to synchronize faster than an eavesdropper using this algorithm.

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

Höche, Stefan; Reichelt, Daniel; Siegert, Frank

We present a systematic study of differences between NLL resummation and parton showers. We first construct a Markovian Monte-Carlo algorithm for resummation of additive observables in electron-positron annihilation. Approximations intrinsic to the pure NLL result are then removed, in order to obtain a traditional, momentum and probability conserving parton shower based on the coherent branching formalism. The impact of each approximation is studied, and an overall comparison is made between the parton shower and pure NLL resummation. Differences compared to modern parton-shower algorithms formulated in terms of color dipoles are analyzed.

On estimating the phase of a periodic waveform in additive Gaussian noise, part 3

NASA Technical Reports Server (NTRS)

Rauch, L. L.

1991-01-01

Motivated by advances in signal processing technology that support more complex algorithms, researchers have taken a new look at the problem of estimating the phase and other parameters of a nearly periodic waveform in additive Gaussian noise, based on observation during a given time interval. Parts 1 and 2 are very briefly reviewed. In part 3, the actual performances of some of the highly nonlinear estimation algorithms of parts 1 and 2 are evaluated by numerical simulation using Monte Carlo techniques.

Determining the Complexity of the Quantum Adiabatic Algorithm using Quantum Monte Carlo Simulations

DTIC Science & Technology

2012-12-18

of this printing. List the papers, including journal references, in the following categories: Received Paper 12/06/2012 4.00 Itay Hen, A. Young...PhysRevLett.104.020502 12/06/2012 3.00 A. P. Young, Itay Hen. Exponential complexity of the quantum adiabatic algorithm for certain satisfiability problems...Physical Review E, (12 2011): 0. doi: 10.1103/PhysRevE.84.061152 12/06/2012 5.00 Edward Farhi, David Gosset, Itay Hen, A. Sandvik, Peter Shor, A

Comparison between extreme learning machine and wavelet neural networks in data classification

NASA Astrophysics Data System (ADS)

Yahia, Siwar; Said, Salwa; Jemai, Olfa; Zaied, Mourad; Ben Amar, Chokri

2017-03-01

Extreme learning Machine is a well known learning algorithm in the field of machine learning. It's about a feed forward neural network with a single-hidden layer. It is an extremely fast learning algorithm with good generalization performance. In this paper, we aim to compare the Extreme learning Machine with wavelet neural networks, which is a very used algorithm. We have used six benchmark data sets to evaluate each technique. These datasets Including Wisconsin Breast Cancer, Glass Identification, Ionosphere, Pima Indians Diabetes, Wine Recognition and Iris Plant. Experimental results have shown that both extreme learning machine and wavelet neural networks have reached good results.

Optimizing the Learning Order of Chinese Characters Using a Novel Topological Sort Algorithm

PubMed Central

Wang, Jinzhao

2016-01-01

We present a novel algorithm for optimizing the order in which Chinese characters are learned, one that incorporates the benefits of learning them in order of usage frequency and in order of their hierarchal structural relationships. We show that our work outperforms previously published orders and algorithms. Our algorithm is applicable to any scheduling task where nodes have intrinsic differences in importance and must be visited in topological order. PMID:27706234

A Configurational-Bias-Monte-Carlo Back-Mapping Algorithm for Efficient and Rapid Conversion of Coarse-Grained Water Structures Into Atomistic Models.

PubMed

Loeffler, Troy David; Chan, Henry; Narayanan, Badri; Cherukara, Mathew J; Gray, Stephen K; Sankaranarayanan, Subramanian K R S

2018-06-20

Coarse-grained molecular dynamics (MD) simulations represent a powerful approach to simulate longer time scale and larger length scale phenomena than those accessible to all-atom models. The gain in efficiency, however, comes at the cost of atomistic details. The reverse transformation, also known as back-mapping, of coarse grained beads into their atomistic constituents represents a major challenge. Most existing approaches are limited to specific molecules or specific force-fields and often rely on running a long time atomistic MD of the back-mapped configuration to arrive at an optimal solution. Such approaches are problematic when dealing with systems with high diffusion barriers. Here, we introduce a new extension of the configurational-bias-Monte-Carlo (CBMC) algorithm, which we term the crystalline-configurational-bias-Monte-Carlo (C-CBMC) algortihm, that allows rapid and efficient conversion of a coarse-grained model back into its atomistic representation. Although the method is generic, we use a coarse-grained water model as a representative example and demonstrate the back-mapping or reverse transformation for model systems ranging from the ice-liquid water interface to amorphous and crystalline ice configurations. A series of simulations using the TIP4P/Ice model are performed to compare the new CBMC method to several other standard Monte Carlo and Molecular Dynamics based back-mapping techniques. In all the cases, the C-CBMC algorithm is able to find optimal hydrogen bonded configuration many thousand evaluations/steps sooner than the other methods compared within this paper. For crystalline ice structures such as a hexagonal, cubic, and cubic-hexagonal stacking disorder structures, the C-CBMC was able to find structures that were between 0.05 and 0.1 eV/water molecule lower in energy than the ground state energies predicted by the other methods. Detailed analysis of the atomistic structures show a significantly better global hydrogen positioning when contrasted with the existing simpler back-mapping methods. Our results demonstrate the efficiency and efficacy of our new back-mapping approach, especially for crystalline systems where simple force-field based relaxations have a tendency to get trapped in local minima.

Determining the Number of Participants Needed for the Usability Evaluation of E-Learning Resources: A Monte Carlo Simulation

ERIC Educational Resources Information Center

Davids, Mogamat Razeen; Harvey, Justin; Halperin, Mitchell L.; Chikte, Usuf M. E.

2015-01-01

The usability of computer interfaces has a major influence on learning. Optimising the usability of e-learning resources is therefore essential. However, this may be neglected because of time and monetary constraints. User testing is a common approach to usability evaluation and involves studying typical end-users interacting with the application…

A study of hydrogen diffusion flames using PDF turbulence model

NASA Technical Reports Server (NTRS)

Hsu, Andrew T.

1991-01-01

The application of probability density function (pdf) turbulence models is addressed. For the purpose of accurate prediction of turbulent combustion, an algorithm that combines a conventional computational fluid dynamic (CFD) flow solver with the Monte Carlo simulation of the pdf evolution equation was developed. The algorithm was validated using experimental data for a heated turbulent plane jet. The study of H2-F2 diffusion flames was carried out using this algorithm. Numerical results compared favorably with experimental data. The computations show that the flame center shifts as the equivalence ratio changes, and that for the same equivalence ratio, similarity solutions for flames exist.

A study of hydrogen diffusion flames using PDF turbulence model

NASA Technical Reports Server (NTRS)

Hsu, Andrew T.

1991-01-01

The application of probability density function (pdf) turbulence models is addressed in this work. For the purpose of accurate prediction of turbulent combustion, an algorithm that combines a conventional CFD flow solver with the Monte Carlo simulation of the pdf evolution equation has been developed. The algorithm has been validated using experimental data for a heated turbulent plane jet. The study of H2-F2 diffusion flames has been carried out using this algorithm. Numerical results compared favorably with experimental data. The computuations show that the flame center shifts as the equivalence ratio changes, and that for the same equivalence ratio, similarity solutions for flames exist.

Torsional anharmonicity in the conformational thermodynamics of flexible molecules

NASA Astrophysics Data System (ADS)

Miller, Thomas F., III; Clary, David C.

We present an algorithm for calculating the conformational thermodynamics of large, flexible molecules that combines ab initio electronic structure theory calculations with a torsional path integral Monte Carlo (TPIMC) simulation. The new algorithm overcomes the previous limitations of the TPIMC method by including the thermodynamic contributions of non-torsional vibrational modes and by affordably incorporating the ab initio calculation of conformer electronic energies, and it improves the conventional ab initio treatment of conformational thermodynamics by accounting for the anharmonicity of the torsional modes. Using previously published ab initio results and new TPIMC calculations, we apply the algorithm to the conformers of the adrenaline molecule.

Creating Engaging Online Learning Material with the JSAV JavaScript Algorithm Visualization Library

ERIC Educational Resources Information Center

Karavirta, Ville; Shaffer, Clifford A.

2016-01-01

Data Structures and Algorithms are a central part of Computer Science. Due to their abstract and dynamic nature, they are a difficult topic to learn for many students. To alleviate these learning difficulties, instructors have turned to algorithm visualizations (AV) and AV systems. Research has shown that especially engaging AVs can have an impact…

Research on intelligent algorithm of electro - hydraulic servo control system

NASA Astrophysics Data System (ADS)

Wang, Yannian; Zhao, Yuhui; Liu, Chengtao

2017-09-01

In order to adapt the nonlinear characteristics of the electro-hydraulic servo control system and the influence of complex interference in the industrial field, using a fuzzy PID switching learning algorithm is proposed and a fuzzy PID switching learning controller is designed and applied in the electro-hydraulic servo controller. The designed controller not only combines the advantages of the fuzzy control and PID control, but also introduces the learning algorithm into the switching function, which makes the learning of the three parameters in the switching function can avoid the instability of the system during the switching between the fuzzy control and PID control algorithms. It also makes the switch between these two control algorithm more smoother than that of the conventional fuzzy PID.

Decision tree and ensemble learning algorithms with their applications in bioinformatics.

PubMed

Che, Dongsheng; Liu, Qi; Rasheed, Khaled; Tao, Xiuping

2011-01-01

Machine learning approaches have wide applications in bioinformatics, and decision tree is one of the successful approaches applied in this field. In this chapter, we briefly review decision tree and related ensemble algorithms and show the successful applications of such approaches on solving biological problems. We hope that by learning the algorithms of decision trees and ensemble classifiers, biologists can get the basic ideas of how machine learning algorithms work. On the other hand, by being exposed to the applications of decision trees and ensemble algorithms in bioinformatics, computer scientists can get better ideas of which bioinformatics topics they may work on in their future research directions. We aim to provide a platform to bridge the gap between biologists and computer scientists.

Using deep neural networks to augment NIF post-shot analysis

NASA Astrophysics Data System (ADS)

Humbird, Kelli; Peterson, Luc; McClarren, Ryan; Field, John; Gaffney, Jim; Kruse, Michael; Nora, Ryan; Spears, Brian

2017-10-01

Post-shot analysis of National Ignition Facility (NIF) experiments is the process of determining which simulation inputs yield results consistent with experimental observations. This analysis is typically accomplished by running suites of manually adjusted simulations, or Monte Carlo sampling surrogate models that approximate the response surfaces of the physics code. These approaches are expensive and often find simulations that match only a small subset of observables simultaneously. We demonstrate an alternative method for performing post-shot analysis using inverse models, which map directly from experimental observables to simulation inputs with quantified uncertainties. The models are created using a novel machine learning algorithm which automates the construction and initialization of deep neural networks to optimize predictive accuracy. We show how these neural networks, trained on large databases of post-shot simulations, can rigorously quantify the agreement between simulation and experiment. This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.

AUTONOMOUS GAUSSIAN DECOMPOSITION

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

Lindner, Robert R.; Vera-Ciro, Carlos; Murray, Claire E.

2015-04-15

We present a new algorithm, named Autonomous Gaussian Decomposition (AGD), for automatically decomposing spectra into Gaussian components. AGD uses derivative spectroscopy and machine learning to provide optimized guesses for the number of Gaussian components in the data, and also their locations, widths, and amplitudes. We test AGD and find that it produces results comparable to human-derived solutions on 21 cm absorption spectra from the 21 cm SPectral line Observations of Neutral Gas with the EVLA (21-SPONGE) survey. We use AGD with Monte Carlo methods to derive the H i line completeness as a function of peak optical depth and velocitymore » width for the 21-SPONGE data, and also show that the results of AGD are stable against varying observational noise intensity. The autonomy and computational efficiency of the method over traditional manual Gaussian fits allow for truly unbiased comparisons between observations and simulations, and for the ability to scale up and interpret the very large data volumes from the upcoming Square Kilometer Array and pathfinder telescopes.« less

Learning Intelligent Genetic Algorithms Using Japanese Nonograms

ERIC Educational Resources Information Center

Tsai, Jinn-Tsong; Chou, Ping-Yi; Fang, Jia-Cen

2012-01-01

An intelligent genetic algorithm (IGA) is proposed to solve Japanese nonograms and is used as a method in a university course to learn evolutionary algorithms. The IGA combines the global exploration capabilities of a canonical genetic algorithm (CGA) with effective condensed encoding, improved fitness function, and modified crossover and…

Location-Aware Mobile Learning of Spatial Algorithms

ERIC Educational Resources Information Center

Karavirta, Ville

2013-01-01

Learning an algorithm--a systematic sequence of operations for solving a problem with given input--is often difficult for students due to the abstract nature of the algorithms and the data they process. To help students understand the behavior of algorithms, a subfield in computing education research has focused on algorithm…

Sequential and Mixed Genetic Algorithm and Learning Automata (SGALA, MGALA) for Feature Selection in QSAR

PubMed Central

MotieGhader, Habib; Gharaghani, Sajjad; Masoudi-Sobhanzadeh, Yosef; Masoudi-Nejad, Ali

2017-01-01

Feature selection is of great importance in Quantitative Structure-Activity Relationship (QSAR) analysis. This problem has been solved using some meta-heuristic algorithms such as GA, PSO, ACO and so on. In this work two novel hybrid meta-heuristic algorithms i.e. Sequential GA and LA (SGALA) and Mixed GA and LA (MGALA), which are based on Genetic algorithm and learning automata for QSAR feature selection are proposed. SGALA algorithm uses advantages of Genetic algorithm and Learning Automata sequentially and the MGALA algorithm uses advantages of Genetic Algorithm and Learning Automata simultaneously. We applied our proposed algorithms to select the minimum possible number of features from three different datasets and also we observed that the MGALA and SGALA algorithms had the best outcome independently and in average compared to other feature selection algorithms. Through comparison of our proposed algorithms, we deduced that the rate of convergence to optimal result in MGALA and SGALA algorithms were better than the rate of GA, ACO, PSO and LA algorithms. In the end, the results of GA, ACO, PSO, LA, SGALA, and MGALA algorithms were applied as the input of LS-SVR model and the results from LS-SVR models showed that the LS-SVR model had more predictive ability with the input from SGALA and MGALA algorithms than the input from all other mentioned algorithms. Therefore, the results have corroborated that not only is the predictive efficiency of proposed algorithms better, but their rate of convergence is also superior to the all other mentioned algorithms. PMID:28979308

Sequential and Mixed Genetic Algorithm and Learning Automata (SGALA, MGALA) for Feature Selection in QSAR.

PubMed

MotieGhader, Habib; Gharaghani, Sajjad; Masoudi-Sobhanzadeh, Yosef; Masoudi-Nejad, Ali

2017-01-01

Feature selection is of great importance in Quantitative Structure-Activity Relationship (QSAR) analysis. This problem has been solved using some meta-heuristic algorithms such as GA, PSO, ACO and so on. In this work two novel hybrid meta-heuristic algorithms i.e. Sequential GA and LA (SGALA) and Mixed GA and LA (MGALA), which are based on Genetic algorithm and learning automata for QSAR feature selection are proposed. SGALA algorithm uses advantages of Genetic algorithm and Learning Automata sequentially and the MGALA algorithm uses advantages of Genetic Algorithm and Learning Automata simultaneously. We applied our proposed algorithms to select the minimum possible number of features from three different datasets and also we observed that the MGALA and SGALA algorithms had the best outcome independently and in average compared to other feature selection algorithms. Through comparison of our proposed algorithms, we deduced that the rate of convergence to optimal result in MGALA and SGALA algorithms were better than the rate of GA, ACO, PSO and LA algorithms. In the end, the results of GA, ACO, PSO, LA, SGALA, and MGALA algorithms were applied as the input of LS-SVR model and the results from LS-SVR models showed that the LS-SVR model had more predictive ability with the input from SGALA and MGALA algorithms than the input from all other mentioned algorithms. Therefore, the results have corroborated that not only is the predictive efficiency of proposed algorithms better, but their rate of convergence is also superior to the all other mentioned algorithms.

Application of Monte Carlo algorithms to the Bayesian analysis of the Cosmic Microwave Background

NASA Technical Reports Server (NTRS)

Jewell, J.; Levin, S.; Anderson, C. H.

2004-01-01

Power spectrum estimation and evaluation of associated errors in the presence of incomplete sky coverage; nonhomogeneous, correlated instrumental noise; and foreground emission are problems of central importance for the extraction of cosmological information from the cosmic microwave background (CMB).

Error Analyses of the North Alabama Lightning Mapping Array (LMA)

NASA Technical Reports Server (NTRS)

Koshak, W. J.; Solokiewicz, R. J.; Blakeslee, R. J.; Goodman, S. J.; Christian, H. J.; Hall, J. M.; Bailey, J. C.; Krider, E. P.; Bateman, M. G.; Boccippio, D. J.

2003-01-01

Two approaches are used to characterize how accurately the North Alabama Lightning Mapping Array (LMA) is able to locate lightning VHF sources in space and in time. The first method uses a Monte Carlo computer simulation to estimate source retrieval errors. The simulation applies a VHF source retrieval algorithm that was recently developed at the NASA-MSFC and that is similar, but not identical to, the standard New Mexico Tech retrieval algorithm. The second method uses a purely theoretical technique (i.e., chi-squared Curvature Matrix theory) to estimate retrieval errors. Both methods assume that the LMA system has an overall rms timing error of 50ns, but all other possible errors (e.g., multiple sources per retrieval attempt) are neglected. The detailed spatial distributions of retrieval errors are provided. Given that the two methods are completely independent of one another, it is shown that they provide remarkably similar results, except that the chi-squared theory produces larger altitude error estimates than the (more realistic) Monte Carlo simulation.

Chemical application of diffusion quantum Monte Carlo

NASA Technical Reports Server (NTRS)

Reynolds, P. J.; Lester, W. A., Jr.

1984-01-01

The diffusion quantum Monte Carlo (QMC) method gives a stochastic solution to the Schroedinger equation. This approach is receiving increasing attention in chemical applications as a result of its high accuracy. However, reducing statistical uncertainty remains a priority because chemical effects are often obtained as small differences of large numbers. As an example, the single-triplet splitting of the energy of the methylene molecule CH sub 2 is given. The QMC algorithm was implemented on the CYBER 205, first as a direct transcription of the algorithm running on the VAX 11/780, and second by explicitly writing vector code for all loops longer than a crossover length C. The speed of the codes relative to one another as a function of C, and relative to the VAX, are discussed. The computational time dependence obtained versus the number of basis functions is discussed and this is compared with that obtained from traditional quantum chemistry codes and that obtained from traditional computer architectures.

New approach for absolute fluence distribution calculations in Monte Carlo simulations of light propagation in turbid media

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

Böcklin, Christoph, E-mail: boecklic@ethz.ch; Baumann, Dirk; Fröhlich, Jürg

A novel way to attain three dimensional fluence rate maps from Monte-Carlo simulations of photon propagation is presented in this work. The propagation of light in a turbid medium is described by the radiative transfer equation and formulated in terms of radiance. For many applications, particularly in biomedical optics, the fluence rate is a more useful quantity and directly derived from the radiance by integrating over all directions. Contrary to the usual way which calculates the fluence rate from absorbed photon power, the fluence rate in this work is directly calculated from the photon packet trajectory. The voxel based algorithmmore » works in arbitrary geometries and material distributions. It is shown that the new algorithm is more efficient and also works in materials with a low or even zero absorption coefficient. The capabilities of the new algorithm are demonstrated on a curved layered structure, where a non-scattering, non-absorbing layer is sandwiched between two highly scattering layers.« less

Auxiliary-field-based trial wave functions in quantum Monte Carlo calculations

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

Chang, Chia -Chen; Rubenstein, Brenda M.; Morales, Miguel A.

2016-12-19

Quantum Monte Carlo (QMC) algorithms have long relied on Jastrow factors to incorporate dynamic correlation into trial wave functions. While Jastrow-type wave functions have been widely employed in real-space algorithms, they have seen limited use in second-quantized QMC methods, particularly in projection methods that involve a stochastic evolution of the wave function in imaginary time. Here we propose a scheme for generating Jastrow-type correlated trial wave functions for auxiliary-field QMC methods. The method is based on decoupling the two-body Jastrow into one-body projectors coupled to auxiliary fields, which then operate on a single determinant to produce a multideterminant trial wavemore » function. We demonstrate that intelligent sampling of the most significant determinants in this expansion can produce compact trial wave functions that reduce errors in the calculated energies. Lastly, our technique may be readily generalized to accommodate a wide range of two-body Jastrow factors and applied to a variety of model and chemical systems.« less

A comparative study of history-based versus vectorized Monte Carlo methods in the GPU/CUDA environment for a simple neutron eigenvalue problem

NASA Astrophysics Data System (ADS)

Liu, Tianyu; Du, Xining; Ji, Wei; Xu, X. George; Brown, Forrest B.

2014-06-01

For nuclear reactor analysis such as the neutron eigenvalue calculations, the time consuming Monte Carlo (MC) simulations can be accelerated by using graphics processing units (GPUs). However, traditional MC methods are often history-based, and their performance on GPUs is affected significantly by the thread divergence problem. In this paper we describe the development of a newly designed event-based vectorized MC algorithm for solving the neutron eigenvalue problem. The code was implemented using NVIDIA's Compute Unified Device Architecture (CUDA), and tested on a NVIDIA Tesla M2090 GPU card. We found that although the vectorized MC algorithm greatly reduces the occurrence of thread divergence thus enhancing the warp execution efficiency, the overall simulation speed is roughly ten times slower than the history-based MC code on GPUs. Profiling results suggest that the slow speed is probably due to the memory access latency caused by the large amount of global memory transactions. Possible solutions to improve the code efficiency are discussed.

Label consistent K-SVD: learning a discriminative dictionary for recognition.

PubMed

Jiang, Zhuolin; Lin, Zhe; Davis, Larry S

2013-11-01

A label consistent K-SVD (LC-KSVD) algorithm to learn a discriminative dictionary for sparse coding is presented. In addition to using class labels of training data, we also associate label information with each dictionary item (columns of the dictionary matrix) to enforce discriminability in sparse codes during the dictionary learning process. More specifically, we introduce a new label consistency constraint called "discriminative sparse-code error" and combine it with the reconstruction error and the classification error to form a unified objective function. The optimal solution is efficiently obtained using the K-SVD algorithm. Our algorithm learns a single overcomplete dictionary and an optimal linear classifier jointly. The incremental dictionary learning algorithm is presented for the situation of limited memory resources. It yields dictionaries so that feature points with the same class labels have similar sparse codes. Experimental results demonstrate that our algorithm outperforms many recently proposed sparse-coding techniques for face, action, scene, and object category recognition under the same learning conditions.

Autonomous learning based on cost assumptions: theoretical studies and experiments in robot control.

PubMed

Ribeiro, C H; Hemerly, E M

2000-02-01

Autonomous learning techniques are based on experience acquisition. In most realistic applications, experience is time-consuming: it implies sensor reading, actuator control and algorithmic update, constrained by the learning system dynamics. The information crudeness upon which classical learning algorithms operate make such problems too difficult and unrealistic. Nonetheless, additional information for facilitating the learning process ideally should be embedded in such a way that the structural, well-studied characteristics of these fundamental algorithms are maintained. We investigate in this article a more general formulation of the Q-learning method that allows for a spreading of information derived from single updates towards a neighbourhood of the instantly visited state and converges to optimality. We show how this new formulation can be used as a mechanism to safely embed prior knowledge about the structure of the state space, and demonstrate it in a modified implementation of a reinforcement learning algorithm in a real robot navigation task.

Machine learning of molecular properties: Locality and active learning

NASA Astrophysics Data System (ADS)

Gubaev, Konstantin; Podryabinkin, Evgeny V.; Shapeev, Alexander V.

2018-06-01

In recent years, the machine learning techniques have shown great potent1ial in various problems from a multitude of disciplines, including materials design and drug discovery. The high computational speed on the one hand and the accuracy comparable to that of density functional theory on another hand make machine learning algorithms efficient for high-throughput screening through chemical and configurational space. However, the machine learning algorithms available in the literature require large training datasets to reach the chemical accuracy and also show large errors for the so-called outliers—the out-of-sample molecules, not well-represented in the training set. In the present paper, we propose a new machine learning algorithm for predicting molecular properties that addresses these two issues: it is based on a local model of interatomic interactions providing high accuracy when trained on relatively small training sets and an active learning algorithm of optimally choosing the training set that significantly reduces the errors for the outliers. We compare our model to the other state-of-the-art algorithms from the literature on the widely used benchmark tests.

CME leaving the Sun [Video

NASA Image and Video Library

2017-12-08

Animation of a CME leaving the Sun, slamming into our magnetosphere. Credit: NASA/GSFC/SOHO/ESA Sound: Juan Carlos Garcia To learn more go to the SOHO website: sohowww.nascom.nasa.gov/home.html To learn more about NASA's Sun Earth Day go here: sunearthday.nasa.gov/2010/index.php

An efficient Cellular Potts Model algorithm that forbids cell fragmentation

NASA Astrophysics Data System (ADS)

Durand, Marc; Guesnet, Etienne

2016-11-01

The Cellular Potts Model (CPM) is a lattice based modeling technique which is widely used for simulating cellular patterns such as foams or biological tissues. Despite its realism and generality, the standard Monte Carlo algorithm used in the scientific literature to evolve this model preserves connectivity of cells on a limited range of simulation temperature only. We present a new algorithm in which cell fragmentation is forbidden for all simulation temperatures. This allows to significantly enhance realism of the simulated patterns. It also increases the computational efficiency compared with the standard CPM algorithm even at same simulation temperature, thanks to the time spared in not doing unrealistic moves. Moreover, our algorithm restores the detailed balance equation, ensuring that the long-term stage is independent of the chosen acceptance rate and chosen path in the temperature space.