Coseismic and aseismic deformations of the rock mass around deep level mining in South Africa - Joint South African and Japanese study

NASA Astrophysics Data System (ADS)

Milev, A. M.; Yabe, Y.; Naoi, M. M.; Nakatani, M.; Durrheim, R. J.; Ogasawara, H.; Scholz, C. H.

2010-12-01

Two underground sites in a deep level gold mine in South Africa were instrumented by the Council for Scientific and Industrial Research (CSIR) with tilt meters and seismic monitors. One of the sites was also instrumented by JApanese-German Underground Acoustic emission Research in South Africa (JAGUARS) with a small network, approx. 40 m span, of eight Acoustic Emission (AE) sensors. The rate of tilt, defined as quasi-static deformations, and the seismic ground motion, defined as dynamic deformations, were analysed in order to understand the rock mass behavior around deep level mining. In addition the high frequency AE events recorded at hypocentral distances of about 50m were analysed. This was the first implementation of high frequency AE events at such a great depth (3300m below the surface). A good correspondence between the dynamic and quasi-static deformations was found. The rate of coseismic and aseismic tilt, as well as seismicity recorded by the mine seismic network, are approximately constant until the daily blasting time, which takes place from about 19:30 until shortly before 21:00. During the blasting time and the subsequent seismic events the coseismic and aseismic tilt shows a rapid increase indicated by a rapid change of the tilt during the seismic event. Much of the quasi-static deformation, however, occurs independently of the seismic events and was described as ‘slow’ or aseismic events. During the monitoring period a seismic event with MW 1.9 (2.1) occurred in the vicinity of the instrumented site. This event was recorded by both the CSIR integrated monitoring system and JAGUARS acoustic emotion network. The tilt changes associated with this event showed a well pronounced after-tilt. More than 21,000 AE aftershocks were located in the first 150 hours after the main event. Using the distribution of the AE events the position of the fault in the source area was successfully delineated. The distribution of the AE events following the main shock was related to after tilt in order to quantify post slip behavior of the source. There was no evidence found for coseismic expansion of the source after the main slip. An attempt to associate the different type of deformations with the various fracture regions and geological structures around the stopes was carried out. A model, was introduced in which the coseismic deformations are associated with the stress regime outside the stope fracture envelope and very often located on existing geological structures, while the aseismic deformations are associated with mobilization of fractures and stress relaxation within the fracture envelope.

Parallel Distributed Processing Theory in the Age of Deep Networks.

PubMed

Bowers, Jeffrey S

2017-12-01

Parallel distributed processing (PDP) models in psychology are the precursors of deep networks used in computer science. However, only PDP models are associated with two core psychological claims, namely that all knowledge is coded in a distributed format and cognition is mediated by non-symbolic computations. These claims have long been debated in cognitive science, and recent work with deep networks speaks to this debate. Specifically, single-unit recordings show that deep networks learn units that respond selectively to meaningful categories, and researchers are finding that deep networks need to be supplemented with symbolic systems to perform some tasks. Given the close links between PDP and deep networks, it is surprising that research with deep networks is challenging PDP theory. Copyright © 2017. Published by Elsevier Ltd.

Thermalnet: a Deep Convolutional Network for Synthetic Thermal Image Generation

NASA Astrophysics Data System (ADS)

Kniaz, V. V.; Gorbatsevich, V. S.; Mizginov, V. A.

2017-05-01

Deep convolutional neural networks have dramatically changed the landscape of the modern computer vision. Nowadays methods based on deep neural networks show the best performance among image recognition and object detection algorithms. While polishing of network architectures received a lot of scholar attention, from the practical point of view the preparation of a large image dataset for a successful training of a neural network became one of major challenges. This challenge is particularly profound for image recognition in wavelengths lying outside the visible spectrum. For example no infrared or radar image datasets large enough for successful training of a deep neural network are available to date in public domain. Recent advances of deep neural networks prove that they are also capable to do arbitrary image transformations such as super-resolution image generation, grayscale image colorisation and imitation of style of a given artist. Thus a natural question arise: how could be deep neural networks used for augmentation of existing large image datasets? This paper is focused on the development of the Thermalnet deep convolutional neural network for augmentation of existing large visible image datasets with synthetic thermal images. The Thermalnet network architecture is inspired by colorisation deep neural networks.

Is Multitask Deep Learning Practical for Pharma?

PubMed

Ramsundar, Bharath; Liu, Bowen; Wu, Zhenqin; Verras, Andreas; Tudor, Matthew; Sheridan, Robert P; Pande, Vijay

2017-08-28

Multitask deep learning has emerged as a powerful tool for computational drug discovery. However, despite a number of preliminary studies, multitask deep networks have yet to be widely deployed in the pharmaceutical and biotech industries. This lack of acceptance stems from both software difficulties and lack of understanding of the robustness of multitask deep networks. Our work aims to resolve both of these barriers to adoption. We introduce a high-quality open-source implementation of multitask deep networks as part of the DeepChem open-source platform. Our implementation enables simple python scripts to construct, fit, and evaluate sophisticated deep models. We use our implementation to analyze the performance of multitask deep networks and related deep models on four collections of pharmaceutical data (three of which have not previously been analyzed in the literature). We split these data sets into train/valid/test using time and neighbor splits to test multitask deep learning performance under challenging conditions. Our results demonstrate that multitask deep networks are surprisingly robust and can offer strong improvement over random forests. Our analysis and open-source implementation in DeepChem provide an argument that multitask deep networks are ready for widespread use in commercial drug discovery.

Working Towards Deep-Ocean Temperature Monitoring by Studying the Acoustic Ambient Noise Field in the South Pacific Ocean

NASA Astrophysics Data System (ADS)

Sambell, K.; Evers, L. G.; Snellen, M.

2017-12-01

Deriving the deep-ocean temperature is a challenge. In-situ observations and satellite observations are hardly applicable. However, knowledge about changes in the deep ocean temperature is important in relation to climate change. Oceans are filled with low-frequency sound waves created by sources such as underwater volcanoes, earthquakes and seismic surveys. The propagation of these sound waves is temperature dependent and therefore carries valuable information that can be used for temperature monitoring. This phenomenon is investigated by applying interferometry to hydroacoustic data measured in the South Pacific Ocean. The data is measured at hydrophone station H03 which is part of the International Monitoring System (IMS). This network consists of several stations around the world and is in place for the verification of the Comprehensive Nuclear-Test-Ban Treaty (CTBT). The station consists of two arrays located north and south of Robinson Crusoe Island separated by 50 km. Both arrays consist of three hydrophones with an intersensor distance of 2 km located at a depth of 1200 m. This depth is in range of the SOFAR channel. Hydroacoustic data measured at the south station is cross-correlated for the time period 2014-2017. The results are improved by applying one-bit normalization as a preprocessing step. Furthermore, beamforming is applied to the hydroacoustic data in order to characterize ambient noise sources around the array. This shows the presence of a continuous source at a backazimuth between 180 and 200 degrees throughout the whole time period, which is in agreement with the results obtained by cross-correlation. Studies on source strength show a seasonal dependence. This is an indication that the sound is related to acoustic activity in Antarctica. Results on this are supported by acoustic propagation modeling. The normal mode technique is used to study the sound propagation from possible source locations towards station H03.

Resolution of Singularities Introduced by Hierarchical Structure in Deep Neural Networks.

PubMed

Nitta, Tohru

2017-10-01

We present a theoretical analysis of singular points of artificial deep neural networks, resulting in providing deep neural network models having no critical points introduced by a hierarchical structure. It is considered that such deep neural network models have good nature for gradient-based optimization. First, we show that there exist a large number of critical points introduced by a hierarchical structure in deep neural networks as straight lines, depending on the number of hidden layers and the number of hidden neurons. Second, we derive a sufficient condition for deep neural networks having no critical points introduced by a hierarchical structure, which can be applied to general deep neural networks. It is also shown that the existence of critical points introduced by a hierarchical structure is determined by the rank and the regularity of weight matrices for a specific class of deep neural networks. Finally, two kinds of implementation methods of the sufficient conditions to have no critical points are provided. One is a learning algorithm that can avoid critical points introduced by the hierarchical structure during learning (called avoidant learning algorithm). The other is a neural network that does not have some critical points introduced by the hierarchical structure as an inherent property (called avoidant neural network).

Deep South Atlantic carbonate chemistry and increased interocean deep water exchange during last deglaciation

NASA Astrophysics Data System (ADS)

Yu, Jimin; Anderson, Robert F.; Jin, Zhangdong; Menviel, Laurie; Zhang, Fei; Ryerson, Fredrick J.; Rohling, Eelco J.

2014-04-01

Carbon release from the deep ocean at glacial terminations is a critical component of past climate change, but the underlying mechanisms remain poorly understood. We present a 28,000-year high-resolution record of carbonate ion concentration, a key parameter of the global carbon cycle, at 5-km water depth in the South Atlantic. We observe similar carbonate ion concentrations between the Last Glacial Maximum and the late Holocene, despite elevated concentrations in the glacial surface ocean. This strongly supports the importance of respiratory carbon accumulation in a stratified deep ocean for atmospheric CO2 reduction during the last ice age. After ˜9 μmol/kg decline during Heinrich Stadial 1, deep South Atlantic carbonate ion concentration rose by ˜24 μmol/kg from the onset of Bølling to Pre-boreal, likely caused by strengthening North Atlantic Deep Water formation (Bølling) or increased ventilation in the Southern Ocean (Younger Drays) or both (Pre-boreal). The ˜15 μmol/kg decline in deep water carbonate ion since ˜10 ka is consistent with extraction of alkalinity from seawater by deep-sea CaCO3 compensation and coral reef growth on continental shelves during the Holocene. Between 16,600 and 15,000 years ago, deep South Atlantic carbonate ion values converged with those at 3.4-km water depth in the western equatorial Pacific, as did carbon isotope and radiocarbon values. These observations suggest a period of enhanced lateral exchange of carbon between the deep South Atlantic and Pacific Oceans, probably due to an increased transfer of momentum from southern westerlies to the Southern Ocean. By spreading carbon-rich deep Pacific waters around Antarctica for upwelling, invigorated interocean deep water exchange would lead to more efficient CO2 degassing from the Southern Ocean, and thus to an atmospheric CO2 rise, during the early deglaciation.

77 FR 53776 - Snapper-Grouper Fishery of the South Atlantic; Accountability Measures and Commercial Closures...

Federal Register 2010, 2011, 2012, 2013, 2014

2012-09-04

... respective annual catch limit (ACLs) for the deep-water complex (including yellowedge grouper, blueline... the snapper-grouper resource. DATES: The closure for the deep-water complex as well as the porgy...-grouper fishery of the South Atlantic, which includes yellowtail snapper, gray triggerfish, the deep-water...

Research on image retrieval using deep convolutional neural network combining L1 regularization and PRelu activation function

NASA Astrophysics Data System (ADS)

QingJie, Wei; WenBin, Wang

2017-06-01

In this paper, the image retrieval using deep convolutional neural network combined with regularization and PRelu activation function is studied, and improves image retrieval accuracy. Deep convolutional neural network can not only simulate the process of human brain to receive and transmit information, but also contains a convolution operation, which is very suitable for processing images. Using deep convolutional neural network is better than direct extraction of image visual features for image retrieval. However, the structure of deep convolutional neural network is complex, and it is easy to over-fitting and reduces the accuracy of image retrieval. In this paper, we combine L1 regularization and PRelu activation function to construct a deep convolutional neural network to prevent over-fitting of the network and improve the accuracy of image retrieval

The Deep Space Network, volume 17

NASA Technical Reports Server (NTRS)

1973-01-01

The objectives, functions, and organization of the Deep Space Network are summarized. The Deep Space Instrumentation Facility, the Ground Communications Facility, and the Network Control System are described.

Countermeasure Study on Deep-sea Oil Exploitation in the South China Sea——A Comparison between Deep-sea Oil Exploitation in the South China Sea and the Gulf of Mexico

NASA Astrophysics Data System (ADS)

Zhao, Hui; Qiu, Weiting; Qu, Weilu

2018-02-01

The unpromising situation of terrestrial oil resources makes the deep-sea oil industry become an important development strategy. The South China Sea has a vast sea area with a wide distribution of oil and gas resources, but there is a phenomenon that exploration and census rates and oil exploitation are low. In order to solve the above problems, this article analyzes the geology, oil and gas exploration and exploration equipment in the South China Sea and the Gulf of Mexico. Comparing the political environment of China and the United States energy industry and the economic environment of oil companies, this article points out China’s deep-sea oil exploration and mining problems that may exist. Finally, the feasibility of oil exploration and exploitation in the South China Sea is put forward, which will provide reference to improve the conditions of oil exploration in the South China Sea and promoting the stable development of China’s oil industry.

Stable architectures for deep neural networks

NASA Astrophysics Data System (ADS)

Haber, Eldad; Ruthotto, Lars

2018-01-01

Deep neural networks have become invaluable tools for supervised machine learning, e.g. classification of text or images. While often offering superior results over traditional techniques and successfully expressing complicated patterns in data, deep architectures are known to be challenging to design and train such that they generalize well to new data. Critical issues with deep architectures are numerical instabilities in derivative-based learning algorithms commonly called exploding or vanishing gradients. In this paper, we propose new forward propagation techniques inspired by systems of ordinary differential equations (ODE) that overcome this challenge and lead to well-posed learning problems for arbitrarily deep networks. The backbone of our approach is our interpretation of deep learning as a parameter estimation problem of nonlinear dynamical systems. Given this formulation, we analyze stability and well-posedness of deep learning and use this new understanding to develop new network architectures. We relate the exploding and vanishing gradient phenomenon to the stability of the discrete ODE and present several strategies for stabilizing deep learning for very deep networks. While our new architectures restrict the solution space, several numerical experiments show their competitiveness with state-of-the-art networks.

Academia Sinica, TW E-science to Assistant Seismic Observations for Earthquake Research, Monitor and Hazard Reduction Surrounding the South China Sea

NASA Astrophysics Data System (ADS)

Huang, Bor-Shouh; Liu, Chun-Chi; Yen, Eric; Liang, Wen-Tzong; Lin, Simon C.; Huang, Win-Gee; Lee, Shiann-Jong; Chen, Hsin-Yen

Experience from the 1994 giant Sumatra earthquake, seismic and tsunami hazard have been considered as important issues in the South China Sea and its surrounding region, and attracted many seismologist's interesting. Currently, more than 25 broadband seismic instruments are currently operated by Institute of Earth Sciences, Academia Sinica in northern Vietnam to study the geodynamic evolution of the Red river fracture zone and rearranged to distribute to southern Vietnam recently to study the geodynamic evolution and its deep structures of the South China Sea. Similar stations are planned to deploy in Philippines in near future. In planning, some high quality stations may be as permanent stations and added continuous GPS observations, and instruments to be maintained and operated by several cooperation institutes, for instance, Institute of Geophysics, Vietnamese Acadamy of Sciences and Technology in Vietnam and Philippine Institute of Volcanology and Seismology in Philippines. Finally, those stations will be planed to upgrade as real time transmission stations for earthquake monitoring and tsunami warning. However, high speed data transfer within different agencies is always a critical issue for successful network operation. By taking advantage of both EGEE and EUAsiaGrid e-Infrastructure, Academia Sinica Grid Computing Centre coordinates researchers from various Asian countries to construct a platform to high performance data transfer for huge parallel computation. Efforts from this data service and a newly build earthquake data centre for data management may greatly improve seismic network performance. Implementation of Grid infrastructure and e-science issues in this region may assistant development of earthquake research, monitor and natural hazard reduction. In the near future, we will search for new cooperation continually from the surrounding countries of the South China Sea to install new seismic stations to construct a complete seismic network of the South China Sea and encourage studies for earthquake sciences and natural hazard reductions.

The Deep Space Network. [tracking and communication functions and facilities

NASA Technical Reports Server (NTRS)

1974-01-01

The objectives, functions, and organization of the Deep Space Network are summarized. The Deep Space Instrumentation Facility, the Ground Communications Facility, and the Network Control System are described.

The deep space network, volume 13

NASA Technical Reports Server (NTRS)

1973-01-01

The objectives, functions, and organization of the Deep Space Network are summarized. The deep space instrumentation facility, the ground communications facility, and the network control system are described. Other areas reported include: Helios Mission support, DSN support of the Mariner Mars 1971 extended mission, Mariner Venus/Mercury 1973 mission support, Viking mission support, radio science, tracking and ground-based navigation, network control and data processing, and deep space stations.

Deep Learning and Developmental Learning: Emergence of Fine-to-Coarse Conceptual Categories at Layers of Deep Belief Network.

PubMed

Sadeghi, Zahra

2016-09-01

In this paper, I investigate conceptual categories derived from developmental processing in a deep neural network. The similarity matrices of deep representation at each layer of neural network are computed and compared with their raw representation. While the clusters generated by raw representation stand at the basic level of abstraction, conceptual categories obtained from deep representation shows a bottom-up transition procedure. Results demonstrate a developmental course of learning from specific to general level of abstraction through learned layers of representations in a deep belief network. © The Author(s) 2016.

Deep resistivity sounding studies in detecting shear zones: A case study from the southern granulite terrain of India

NASA Astrophysics Data System (ADS)

Singh, S. B.; Stephen, Jimmy

2006-10-01

The resistivity signatures of the major crustal scale shear zones that dissect the southern granulite terrain (SGT) of South India into discrete geological fragments have been investigated. Resistivity structures deduced from deep resistivity sounding measurements acquired with a 10 km long Schlumberger spreads yield significant insights into the resistivity distribution within the E-W trending shear system comprising the Moyar-Bhavani-Salem-Attur shear zone (MBSASZ) and Palghat-Cauvery shear zone (PCSZ). Vertical and lateral extensions of low resistivity features indicate the possible existence of weak zones at different depths throughout the shear zones. The MBSASZ characterized by very low resistivity in its deeper parts (>2500 m), extends towards the south with slightly higher resistivities to encompass the PCSZ. A major resistivity transition between the northern and southern parts is evident in the two-dimensional resistivity images. The northern Archaean granulite terrain exhibits a higher resistivity than the southern Neoproterozoic granulite terrain. Though this resistivity transition is not clear at greater depths, the extension of low resistivity zones has been well manifested. It is speculated here that a network of crustal scale shear zones in the SGT may have influenced the strength of the lithosphere.

Deep-sea geohazards in the South China Sea

NASA Astrophysics Data System (ADS)

Wu, Shiguo; Wang, Dawei; Völker, David

2018-02-01

Various geological processes and features that might inflict hazards identified in the South China Sea by using new technologies and methods. These features include submarine landslides, pockmark fields, shallow free gas, gas hydrates, mud diapirs and earthquake tsunami, which are widely distributed in the continental slope and reefal islands of the South China Sea. Although the study and assessment of geohazards in the South China Sea came into operation only recently, advances in various aspects are evolving at full speed to comply with National Marine Strategy and `the Belt and Road' Policy. The characteristics of geohazards in deep-water seafloor of the South China Sea are summarized based on new scientific advances. This progress is aimed to aid ongoing deep-water drilling activities and decrease geological risks in ocean development.

Deep learning for computational chemistry.

PubMed

Goh, Garrett B; Hodas, Nathan O; Vishnu, Abhinav

2017-06-15

The rise and fall of artificial neural networks is well documented in the scientific literature of both computer science and computational chemistry. Yet almost two decades later, we are now seeing a resurgence of interest in deep learning, a machine learning algorithm based on multilayer neural networks. Within the last few years, we have seen the transformative impact of deep learning in many domains, particularly in speech recognition and computer vision, to the extent that the majority of expert practitioners in those field are now regularly eschewing prior established models in favor of deep learning models. In this review, we provide an introductory overview into the theory of deep neural networks and their unique properties that distinguish them from traditional machine learning algorithms used in cheminformatics. By providing an overview of the variety of emerging applications of deep neural networks, we highlight its ubiquity and broad applicability to a wide range of challenges in the field, including quantitative structure activity relationship, virtual screening, protein structure prediction, quantum chemistry, materials design, and property prediction. In reviewing the performance of deep neural networks, we observed a consistent outperformance against non-neural networks state-of-the-art models across disparate research topics, and deep neural network-based models often exceeded the "glass ceiling" expectations of their respective tasks. Coupled with the maturity of GPU-accelerated computing for training deep neural networks and the exponential growth of chemical data on which to train these networks on, we anticipate that deep learning algorithms will be a valuable tool for computational chemistry. © 2017 Wiley Periodicals, Inc. © 2017 Wiley Periodicals, Inc.

Deep learning for computational chemistry

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

Goh, Garrett B.; Hodas, Nathan O.; Vishnu, Abhinav

The rise and fall of artificial neural networks is well documented in the scientific literature of both the fields of computer science and computational chemistry. Yet almost two decades later, we are now seeing a resurgence of interest in deep learning, a machine learning algorithm based on “deep” neural networks. Within the last few years, we have seen the transformative impact of deep learning the computer science domain, notably in speech recognition and computer vision, to the extent that the majority of practitioners in those field are now regularly eschewing prior established models in favor of deep learning models. Inmore » this review, we provide an introductory overview into the theory of deep neural networks and their unique properties as compared to traditional machine learning algorithms used in cheminformatics. By providing an overview of the variety of emerging applications of deep neural networks, we highlight its ubiquity and broad applicability to a wide range of challenges in the field, including QSAR, virtual screening, protein structure modeling, QM calculations, materials synthesis and property prediction. In reviewing the performance of deep neural networks, we observed a consistent outperformance against non neural networks state-of-the-art models across disparate research topics, and deep neural network based models often exceeded the “glass ceiling” expectations of their respective tasks. Coupled with the maturity of GPU-accelerated computing for training deep neural networks and the exponential growth of chemical data on which to train these networks on, we anticipate that deep learning algorithms will be a useful tool and may grow into a pivotal role for various challenges in the computational chemistry field.« less

A Deep Learning Network Approach to ab initio Protein Secondary Structure Prediction

PubMed Central

Spencer, Matt; Eickholt, Jesse; Cheng, Jianlin

2014-01-01

Ab initio protein secondary structure (SS) predictions are utilized to generate tertiary structure predictions, which are increasingly demanded due to the rapid discovery of proteins. Although recent developments have slightly exceeded previous methods of SS prediction, accuracy has stagnated around 80% and many wonder if prediction cannot be advanced beyond this ceiling. Disciplines that have traditionally employed neural networks are experimenting with novel deep learning techniques in attempts to stimulate progress. Since neural networks have historically played an important role in SS prediction, we wanted to determine whether deep learning could contribute to the advancement of this field as well. We developed an SS predictor that makes use of the position-specific scoring matrix generated by PSI-BLAST and deep learning network architectures, which we call DNSS. Graphical processing units and CUDA software optimize the deep network architecture and efficiently train the deep networks. Optimal parameters for the training process were determined, and a workflow comprising three separately trained deep networks was constructed in order to make refined predictions. This deep learning network approach was used to predict SS for a fully independent test data set of 198 proteins, achieving a Q3 accuracy of 80.7% and a Sov accuracy of 74.2%. PMID:25750595

A Deep Learning Network Approach to ab initio Protein Secondary Structure Prediction.

PubMed

Spencer, Matt; Eickholt, Jesse; Jianlin Cheng

2015-01-01

Ab initio protein secondary structure (SS) predictions are utilized to generate tertiary structure predictions, which are increasingly demanded due to the rapid discovery of proteins. Although recent developments have slightly exceeded previous methods of SS prediction, accuracy has stagnated around 80 percent and many wonder if prediction cannot be advanced beyond this ceiling. Disciplines that have traditionally employed neural networks are experimenting with novel deep learning techniques in attempts to stimulate progress. Since neural networks have historically played an important role in SS prediction, we wanted to determine whether deep learning could contribute to the advancement of this field as well. We developed an SS predictor that makes use of the position-specific scoring matrix generated by PSI-BLAST and deep learning network architectures, which we call DNSS. Graphical processing units and CUDA software optimize the deep network architecture and efficiently train the deep networks. Optimal parameters for the training process were determined, and a workflow comprising three separately trained deep networks was constructed in order to make refined predictions. This deep learning network approach was used to predict SS for a fully independent test dataset of 198 proteins, achieving a Q3 accuracy of 80.7 percent and a Sov accuracy of 74.2 percent.

Deep space network energy program

NASA Technical Reports Server (NTRS)

Friesema, S. E.

1980-01-01

If the Deep Space Network is to exist in a cost effective and reliable manner in the next decade, the problems presented by international energy cost increases and energy availability must be addressed. The Deep Space Network Energy Program was established to implement solutions compatible with the ongoing development of the total network.

Interplanetary CubeSat Navigational Challenges

NASA Technical Reports Server (NTRS)

Martin-Mur, Tomas J.; Gustafson, Eric D.; Young, Brian T.

2015-01-01

CubeSats are miniaturized spacecraft of small mass that comply with a form specification so they can be launched using standardized deployers. Since the launch of the first CubeSat into Earth orbit in June of 2003, hundreds have been placed into orbit. There are currently a number of proposals to launch and operate CubeSats in deep space, including MarCO, a technology demonstration that will launch two CubeSats towards Mars using the same launch vehicle as NASA's Interior Exploration using Seismic Investigations, Geodesy and Heat Transport (InSight) Mars lander mission. The MarCO CubeSats are designed to relay the information transmitted by the InSight UHF radio during Entry, Descent, and Landing (EDL) in real time to the antennas of the Deep Space Network (DSN) on Earth. Other CubeSatts proposals intend to demonstrate the operation of small probes in deep space, investigate the lunar South Pole, and visit a near Earth object, among others. Placing a CubeSat into an interplanetary trajectory makes it even more challenging to pack the necessary power, communications, and navigation capabilities into such a small spacecraft. This paper presents some of the challenges and approaches for successfully navigating CubeSats and other small spacecraft in deep space.

The clementine bistatic radar experiment: Evidence for ice on the moon

USGS Publications Warehouse

Spudis, P.D.; Nozette, S.; Lichtenberg, C.; Bonner, R.; Ort, W.; Malaret, E.; Robinson, M.; Shoemaker, E.

1998-01-01

Ice deposits, derived from comets and water-bearing meteorites hitting the Moon over geological times, have long been postulated to exist in dark areas near the poles of the Moon. The characteristics of radio waves beamed from the Clementine spacecraft into the polar areas, reflected from the Moon's surface, and received on the large dish antennas of the Deep Space Network here on Earth show that roughly the volume of a small lake (???0.9-1.8 km3) of water ice makes up part of the Moon's surface layer near the south pole. The discovery of ice near the lunar south pole has important ramifications for a permanent return to the Moon. These deposits could be used to manufacture rocket propellant and to support human life on the Moon. ?? 1998 MAHK Hayka/Interperiodica Publishing.

The Role of Deep Creep in the Timing of Large Earthquakes

NASA Astrophysics Data System (ADS)

Sammis, C. G.; Smith, S. W.

2012-12-01

The observed temporal clustering of the world's largest earthquakes has been largely discounted for two reasons: a) it is consistent with Poisson clustering, and b) no physical mechanism leading to such clustering has been proposed. This lack of a mechanism arises primarily because the static stress transfer mechanism, commonly used to explain aftershocks and the clustering of large events on localized fault networks, does not work at global distances. However, there is recent observational evidence that the surface waves from large earthquakes trigger non-volcanic tremor at the base of distant fault zones at global distances. Based on these observations, we develop a simple non-linear coupled oscillator model that shows how the triggering of such tremor can lead to the synchronization of large earthquakes on a global scale. A basic assumption of the model is that induced tremor is a proxy for deep creep that advances the seismic cycle of the fault. We support this hypothesis by demonstrating that the 2010 Maule Chile and the 2011 Fukushima Japan earthquakes, which have been shown to induce tremor on the Parkfield segment of the San Andreas Fault, also produce changes in off-fault seismicity that are spatially and temporally consistent with episodes of deep creep on the fault. The observed spatial pattern can be simulated using an Okada dislocation model for deep creep (below 20 km) on the fault plane in which the slip rate decreases from North to South consistent with surface creep measurements and deepens south of the "Parkfield asperity" as indicated by recent tremor locations. The model predicts the off-fault events should have reverse mechanism consistent with observed topography.

Considering the Role of Stress in Populations of High-Risk, Underserved Community Networks Program Centers.

PubMed

Hébert, James R; Braun, Kathryn L; Kaholokula, Joseph Keawe'aimoku; Armstead, Cheryl A; Burch, James B; Thompson, Beti

2015-01-01

Cancer disparities are associated with a broad range of sociocultural determinants of health that operate in community contexts. High-risk populations may be more vulnerable to social and environmental factors that lead to chronic stress. Theoretical and empirical research indicates that exposure to contextual and sociocultural stress alters biological systems, thereby influencing cancer risk, progression, and, ultimately, mortality. We sought to describe contextual pathways through which stress likely increases cancer risk in high-risk, underserved populations. This review presents a description of the link between contextual stressors and disease risk disparities within underserved communities, with a focus on 1) stress as a proximal link between biological processes, such as cytokine responses, inflammation, and cancer and 2) stress as a distal link to cancer through biobehavioral risk factors such as poor diet, physical inactivity, circadian rhythm or sleep disruption, and substance abuse. These concepts are illustrated through application to populations served by three National Cancer Institute-funded Community Networks Program Centers (CNPCs): African Americans in the Deep South (the South Carolina Cancer Disparities Community Network [SCCDCN]), Native Hawaiians ('Imi Hale-Native Hawaiian Cancer Network), and Latinos in the Lower Yakima Valley of Washington State (The Center for Hispanic Health Promotion: Reducing Cancer Disparities). Stress experienced by the underserved communities represented in the CNPCs is marked by social, biological, and behavioral pathways that increase cancer risk. A case is presented to increase research on sociocultural determinants of health, stress, and cancer risk among racial/ethnic minorities in underserved communities.

The Deep Space Network

NASA Technical Reports Server (NTRS)

1979-01-01

Deep Space Network progress in flight project support, tracking and data acquisition, research and technology, network engineering, hardware and software implementation, and operations is cited. Topics covered include: tracking and ground based navigation; spacecraft/ground communication; station control and operations technology; ground communications; and deep space stations.

DSN test and training system

NASA Technical Reports Server (NTRS)

Thorman, H. C.

1975-01-01

Key characteristics of the Deep Space Network Test and Training System were presented. Completion of the Mark III-75 system implementation is reported. Plans are summarized for upgrading the system to a Mark III-77 configuration to support Deep Space Network preparations for the Mariner Jupiter/Saturn 1977 and Pioneer Venus 1978 missions. A general description of the Deep Space Station, Ground Communications Facility, and Network Operations Control Center functions that comprise the Deep Space Network Test and Training System is also presented.

Development and application of deep convolutional neural network in target detection

NASA Astrophysics Data System (ADS)

Jiang, Xiaowei; Wang, Chunping; Fu, Qiang

2018-04-01

With the development of big data and algorithms, deep convolution neural networks with more hidden layers have more powerful feature learning and feature expression ability than traditional machine learning methods, making artificial intelligence surpass human level in many fields. This paper first reviews the development and application of deep convolutional neural networks in the field of object detection in recent years, then briefly summarizes and ponders some existing problems in the current research, and the future development of deep convolutional neural network is prospected.

Waveform Tomography of the South Atlantic Region

NASA Astrophysics Data System (ADS)

Celli, N. L.; Lebedev, S.; Schaeffer, A. J.; Gaina, C.

2016-12-01

The rapid growth in broadband seismic data, along with developments in waveform tomography techniques, allow us to greatly improve the data sampling in the southern hemisphere and resolve the upper-mantle structure beneath the South Atlantic region at a new level of detail. We have gathered a very large waveform dataset, including all publicly available data from permanent and temporary networks. Our S-velocity tomographic model is constrained by vertical-component waveform fits, computed using the Automated Multimode Inversion of surface, S and multiple S waves. Each seismogram fit provides a set of linear equations describing 1D average velocity perturbations within approximate sensitivity volumes, with respect to a 3D reference model. All the equations are then combined into a large linear system and inverted jointly for a model of shear- and compressional-wave speeds and azimuthal anisotropy within the lithosphere and underlying mantle. The isotropic-average shear speeds are proxies for temperature and composition at depth, while azimuthal anisotropy provides evidence on the past and present deformation in the lithosphere and asthenosphere beneath the region. We resolve the complex boundaries of the mantle roots of South America's and Africa's cratons and the deep low-velocity anomalies beneath volcanic areas in South America. Pronounced lithospheric high seismic velocity anomalies beneath the Argentine Basin suggest that its anomalously deep seafloor, previously attributed to dynamic topography, is mainly due to anomalously cold, thick lithosphere. Major hotspots show low-velocity anomalies extending substantially deeper than those beneath the mid-ocean ridge. The Vema Hotspot shows a major, hot asthenospheric anomaly beneath thick, cold oceanic lithosphere. The mantle lithosphere beneath the Walvis Ridge—a hotspot track—shows normal cooling. The volcanic Cameroon Line, in contrast, is characterized by thin lithosphere beneath the locations of recent volcanism.

The deep space network

NASA Technical Reports Server (NTRS)

1974-01-01

The progress is reported of Deep Space Network (DSN) research in the following areas: (1) flight project support, (2) spacecraft/ground communications, (3) station control and operations technology, (4) network control and processing, and (5) deep space stations. A description of the DSN functions and facilities is included.

The Deep Space Network. An instrument for radio navigation of deep space probes

NASA Technical Reports Server (NTRS)

Renzetti, N. A.; Jordan, J. F.; Berman, A. L.; Wackley, J. A.; Yunck, T. P.

1982-01-01

The Deep Space Network (DSN) network configurations used to generate the navigation observables and the basic process of deep space spacecraft navigation, from data generation through flight path determination and correction are described. Special emphasis is placed on the DSN Systems which generate the navigation data: the DSN Tracking and VLBI Systems. In addition, auxiliary navigational support functions are described.

Total Mercury and Methylmercury Response in Water, Sediment, and Biota to Destratification of the Great Salt Lake, Utah, United States.

PubMed

Valdes, Carla; Black, Frank J; Stringham, Blair; Collins, Jeffrey N; Goodman, James R; Saxton, Heidi J; Mansfield, Christopher R; Schmidt, Joshua N; Yang, Shu; Johnson, William P

2017-05-02

Measurements of chemical and physical parameters made before and after sealing of culverts in the railroad causeway spanning the Great Salt Lake in late 2013 documented dramatic alterations in the system in response to the elimination of flow between the Great Salt Lake's north and south arms. The flow of denser, more-saline water through the culverts from the north arm (Gunnison Bay) to the south arm (Gilbert Bay) previously drove the perennial stratification of the south arm and the existence of oxic shallow brine and anoxic deep brine layers. Closure of the causeway culverts occurred concurrently with a multiyear drought that resulted in a decrease in the lake elevation and a concomitant increase in top-down erosion of the upper surface of the deep brine layer by wind-forced mixing. The combination of these events resulted in the replacement of the formerly stratified water column in the south arm with one that was vertically homogeneous and oxic. Total mercury concentrations in the deep waters of the south arm decreased by approximately 81% and methylmercury concentrations in deep waters decreased by roughly 86% due to destratification. Methylmercury concentrations decreased by 77% in underlying surficial sediment, whereas there was no change observed in total mercury. The dramatic mercury loss from deep waters and methylmercury loss from underlying sediment in response to causeway sealing provides new understanding of the potential role of the deep brine layer in the accumulation and persistence of methylmercury in the Great Salt Lake. Additional mercury measurements in biota appear to contradict the previously implied connection between elevated methylmercury concentrations in the deep brine layer and elevated mercury in avian species reported prior to causeway sealing.

Two-Stage Approach to Image Classification by Deep Neural Networks

NASA Astrophysics Data System (ADS)

Ososkov, Gennady; Goncharov, Pavel

2018-02-01

The paper demonstrates the advantages of the deep learning networks over the ordinary neural networks on their comparative applications to image classifying. An autoassociative neural network is used as a standalone autoencoder for prior extraction of the most informative features of the input data for neural networks to be compared further as classifiers. The main efforts to deal with deep learning networks are spent for a quite painstaking work of optimizing the structures of those networks and their components, as activation functions, weights, as well as the procedures of minimizing their loss function to improve their performances and speed up their learning time. It is also shown that the deep autoencoders develop the remarkable ability for denoising images after being specially trained. Convolutional Neural Networks are also used to solve a quite actual problem of protein genetics on the example of the durum wheat classification. Results of our comparative study demonstrate the undoubted advantage of the deep networks, as well as the denoising power of the autoencoders. In our work we use both GPU and cloud services to speed up the calculations.

Taxonomic research on deep-sea macrofauna in the South China Sea using the Chinese deep-sea submersible Jiaolong.

PubMed

Li, Xinzheng

2017-07-01

This paper reviews the taxonomic and biodiversity studies of deep-sea invertebrates in the South China Sea based on the samples collected by the Chinese manned deep-sea submersible Jiaolong. To date, 6 new species have been described, including the sponges Lophophysema eversa, Saccocalyx microhexactin and Semperella jiaolongae as well as the crustaceans Uroptychus jiaolongae, Uroptychus spinulosus and Globospongicola jiaolongi; some newly recorded species from the South China Sea have also been reported. The Bathymodiolus platifrons-Shinkaia crosnieri deep-sea cold seep community has been reported by Li (2015), as has the mitochondrial genome of the glass sponge L. eversa by Zhang et al. (2016). The population structures of two dominant species, the shrimp Shinkaia crosnieri and the mussel Bathymodiolus platifrons, from the cold seep Bathymodiolus platifrons-Shinkaia crosnieri community in the South China Sea and the hydrothermal vents in the Okinawa Trough, were compared using molecular analysis. The systematic position of the shrimp genus Globospongicola was discussed based on 16S rRNA gene sequences. © 2017 International Society of Zoological Sciences, Institute of Zoology/Chinese Academy of Sciences and John Wiley & Sons Australia, Ltd.

Facial expression recognition based on improved deep belief networks

NASA Astrophysics Data System (ADS)

Wu, Yao; Qiu, Weigen

2017-08-01

In order to improve the robustness of facial expression recognition, a method of face expression recognition based on Local Binary Pattern (LBP) combined with improved deep belief networks (DBNs) is proposed. This method uses LBP to extract the feature, and then uses the improved deep belief networks as the detector and classifier to extract the LBP feature. The combination of LBP and improved deep belief networks is realized in facial expression recognition. In the JAFFE (Japanese Female Facial Expression) database on the recognition rate has improved significantly.

The deep space network, volume 7

NASA Technical Reports Server (NTRS)

1972-01-01

The objectives, functions, and organization of the Deep Space Network are summarized. The Deep Space Instrumentation Facility, the Ground Communications Facility, and the Space Flight Operations Facility are described.

Deep Learning Neural Networks and Bayesian Neural Networks in Data Analysis

NASA Astrophysics Data System (ADS)

Chernoded, Andrey; Dudko, Lev; Myagkov, Igor; Volkov, Petr

2017-10-01

Most of the modern analyses in high energy physics use signal-versus-background classification techniques of machine learning methods and neural networks in particular. Deep learning neural network is the most promising modern technique to separate signal and background and now days can be widely and successfully implemented as a part of physical analysis. In this article we compare Deep learning and Bayesian neural networks application as a classifiers in an instance of top quark analysis.

The deep space network

NASA Technical Reports Server (NTRS)

1977-01-01

Presented is Deep Space Network (DSN) progress in flight project support, tracking and data acquisition (TDA) research and technology, network engineering, hardware and software implementation, and operations.

The deep space network

NASA Technical Reports Server (NTRS)

1975-01-01

Summaries are given of Deep Space Network progress in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations.

Lightning Enhancement Over Major Shipping Lanes

NASA Astrophysics Data System (ADS)

Thornton, J. A.; Holzworth, R. H., II; Virts, K.; Mitchell, T. P.

2017-12-01

Using twelve years of high resolution global lightning stroke data from the World Wide Lightning Location Network (WWLLN), we show that lightning density is enhanced by up to a factor of two directly over shipping lanes in the northeastern Indian Ocean and the South China Sea as compared to adjacent areas with similar climatological characteristics. The lightning enhancement is most prominent during the convectively active season, November-April for the Indian Ocean and April - December in the South China Sea, and has been detectable from at least 2005 to the present. We hypothesize that emissions of aerosol particles and precursors by maritime vessel traffic leads to a microphysical enhancement of convection and storm electrification in the region of the shipping lanes. These persistent localized anthropogenic perturbations to otherwise clean regions are a unique opportunity to more thoroughly understand the sensitivity of maritime deep convection and lightning to aerosol particles.

Deep-sea ostracods from the South Atlantic sector of the Southern ocean during the Last 370,000 years

USGS Publications Warehouse

Yasuhara, Moriaki; Cronin, T. M.; Hunt, G.; Hodell, D.A.

2009-01-01

We report changes of deep-sea ostracod fauna during the last 370,000 yr from the Ocean Drilling Program (ODP) Hole 704A in the South Atlantic sector of the Southern Ocean. The results show that faunal changes are coincident with glacial/interglacial-scale deep-water circulation changes, even though our dataset is relatively small and the waters are barren of ostracods until mid-MIS (Marine Isotope Stage) 5. Krithe and Poseidonamicus were dominant during the Holocene interglacial period and the latter part of MIS 5, when this site was under the influence of North Atlantic Deep Water (NADW). Conversely, Henryhowella and Legitimocythere were dominant during glacial periods, when this site was in the path of Circumpolar Deep Water (CPDW). Three new species (Aversovalva brandaoae, Poseidonamicus hisayoae, and Krithe mazziniae) are described herein. This is the first report of Quaternary glacial/interglacial scale deep-sea ostracod faunal changes in the Southern and South Atlantic Oceans, a key region for understanding Quaternary climate and deep-water circulation, although the paucity of Quaternary ostracods in this region necessitates further research. ?? 2009 The Paleontological Society.

Constructing fine-granularity functional brain network atlases via deep convolutional autoencoder.

PubMed

Zhao, Yu; Dong, Qinglin; Chen, Hanbo; Iraji, Armin; Li, Yujie; Makkie, Milad; Kou, Zhifeng; Liu, Tianming

2017-12-01

State-of-the-art functional brain network reconstruction methods such as independent component analysis (ICA) or sparse coding of whole-brain fMRI data can effectively infer many thousands of volumetric brain network maps from a large number of human brains. However, due to the variability of individual brain networks and the large scale of such networks needed for statistically meaningful group-level analysis, it is still a challenging and open problem to derive group-wise common networks as network atlases. Inspired by the superior spatial pattern description ability of the deep convolutional neural networks (CNNs), a novel deep 3D convolutional autoencoder (CAE) network is designed here to extract spatial brain network features effectively, based on which an Apache Spark enabled computational framework is developed for fast clustering of larger number of network maps into fine-granularity atlases. To evaluate this framework, 10 resting state networks (RSNs) were manually labeled from the sparsely decomposed networks of Human Connectome Project (HCP) fMRI data and 5275 network training samples were obtained, in total. Then the deep CAE models are trained by these functional networks' spatial maps, and the learned features are used to refine the original 10 RSNs into 17 network atlases that possess fine-granularity functional network patterns. Interestingly, it turned out that some manually mislabeled outliers in training networks can be corrected by the deep CAE derived features. More importantly, fine granularities of networks can be identified and they reveal unique network patterns specific to different brain task states. By further applying this method to a dataset of mild traumatic brain injury study, it shows that the technique can effectively identify abnormal small networks in brain injury patients in comparison with controls. In general, our work presents a promising deep learning and big data analysis solution for modeling functional connectomes, with fine granularities, based on fMRI data. Copyright © 2017 Elsevier B.V. All rights reserved.

Generation of sedimentary fabrics and facies by repetitive excavation and storm infilling of burrow networks Holocene of south Florida and Caicos Platform, B. W. I

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

Tedesco, L.P.; Wanless, H.R.

Excavation of deep, open burrow networks and subsequent infilling with sediment from the surface produces an entirely new sedimentary deposit and results in obliteration to severe diagenetic transformation of precursor depositional facies. Repetitive excavation and infilling is responsible for creating the preserved depositional facies of many marine deposits. Excavating burrowers occur from intertidal to abyssal depths, and are important throughout the Phanerozoic. The repetitive coupling of deep, open burrow excavation with subsequent storm sediment infilling of open burrow networks is a gradual process that ultimately results in the loss of the original deposit and the generation of new lithologies, fabricsmore » and facies. The new lithologies are produced in the subsurface and possess fabrics, textures and skeletal assemblages that are not a direct reflection of either precursor facies or the surficial depositional conditions. Sedimentary facies generated by repetitive burrow excavation and infilling commonly are massively bedded and generally are mottled skeletal packstones. Skeletal grains usually are well-preserved and coarser components are concentrated in burrow networks, pockets and patches. The coarse skeletal components of burrow-generated facies are a mixture of coarse bioclasts from the precursor facies and both the coarse and fine skeletal material introduced from the sediment surface. Many so-called bioturbated or massive facies may, in fact, be primary depositional facies generated in the subsurface and represent severe diagenetic transformation of originally deposited sequences. In addition, mudstones and wackestones mottled with packstone patches may record storm sedimentation.« less

Wishart Deep Stacking Network for Fast POLSAR Image Classification.

PubMed

Jiao, Licheng; Liu, Fang

2016-05-11

Inspired by the popular deep learning architecture - Deep Stacking Network (DSN), a specific deep model for polarimetric synthetic aperture radar (POLSAR) image classification is proposed in this paper, which is named as Wishart Deep Stacking Network (W-DSN). First of all, a fast implementation of Wishart distance is achieved by a special linear transformation, which speeds up the classification of POLSAR image and makes it possible to use this polarimetric information in the following Neural Network (NN). Then a single-hidden-layer neural network based on the fast Wishart distance is defined for POLSAR image classification, which is named as Wishart Network (WN) and improves the classification accuracy. Finally, a multi-layer neural network is formed by stacking WNs, which is in fact the proposed deep learning architecture W-DSN for POLSAR image classification and improves the classification accuracy further. In addition, the structure of WN can be expanded in a straightforward way by adding hidden units if necessary, as well as the structure of the W-DSN. As a preliminary exploration on formulating specific deep learning architecture for POLSAR image classification, the proposed methods may establish a simple but clever connection between POLSAR image interpretation and deep learning. The experiment results tested on real POLSAR image show that the fast implementation of Wishart distance is very efficient (a POLSAR image with 768000 pixels can be classified in 0.53s), and both the single-hidden-layer architecture WN and the deep learning architecture W-DSN for POLSAR image classification perform well and work efficiently.

deepNF: Deep network fusion for protein function prediction.

PubMed

Gligorijevic, Vladimir; Barot, Meet; Bonneau, Richard

2018-06-01

The prevalence of high-throughput experimental methods has resulted in an abundance of large-scale molecular and functional interaction networks. The connectivity of these networks provides a rich source of information for inferring functional annotations for genes and proteins. An important challenge has been to develop methods for combining these heterogeneous networks to extract useful protein feature representations for function prediction. Most of the existing approaches for network integration use shallow models that encounter difficulty in capturing complex and highly-nonlinear network structures. Thus, we propose deepNF, a network fusion method based on Multimodal Deep Autoencoders to extract high-level features of proteins from multiple heterogeneous interaction networks. We apply this method to combine STRING networks to construct a common low-dimensional representation containing high-level protein features. We use separate layers for different network types in the early stages of the multimodal autoencoder, later connecting all the layers into a single bottleneck layer from which we extract features to predict protein function. We compare the cross-validation and temporal holdout predictive performance of our method with state-of-the-art methods, including the recently proposed method Mashup. Our results show that our method outperforms previous methods for both human and yeast STRING networks. We also show substantial improvement in the performance of our method in predicting GO terms of varying type and specificity. deepNF is freely available at: https://github.com/VGligorijevic/deepNF. vgligorijevic@flatironinstitute.org, rb133@nyu.edu. Supplementary data are available at Bioinformatics online.

Deep Belief Networks for Electroencephalography: A Review of Recent Contributions and Future Outlooks.

PubMed

Movahedi, Faezeh; Coyle, James L; Sejdic, Ervin

2018-05-01

Deep learning, a relatively new branch of machine learning, has been investigated for use in a variety of biomedical applications. Deep learning algorithms have been used to analyze different physiological signals and gain a better understanding of human physiology for automated diagnosis of abnormal conditions. In this paper, we provide an overview of deep learning approaches with a focus on deep belief networks in electroencephalography applications. We investigate the state-of-the-art algorithms for deep belief networks and then cover the application of these algorithms and their performances in electroencephalographic applications. We covered various applications of electroencephalography in medicine, including emotion recognition, sleep stage classification, and seizure detection, in order to understand how deep learning algorithms could be modified to better suit the tasks desired. This review is intended to provide researchers with a broad overview of the currently existing deep belief network methodology for electroencephalography signals, as well as to highlight potential challenges for future research.

The deep space network

NASA Technical Reports Server (NTRS)

1977-01-01

A Deep Space Network progress report is presented dealing with in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations.

Intrusion Detection System Using Deep Neural Network for In-Vehicle Network Security.

PubMed

Kang, Min-Joo; Kang, Je-Won

2016-01-01

A novel intrusion detection system (IDS) using a deep neural network (DNN) is proposed to enhance the security of in-vehicular network. The parameters building the DNN structure are trained with probability-based feature vectors that are extracted from the in-vehicular network packets. For a given packet, the DNN provides the probability of each class discriminating normal and attack packets, and, thus the sensor can identify any malicious attack to the vehicle. As compared to the traditional artificial neural network applied to the IDS, the proposed technique adopts recent advances in deep learning studies such as initializing the parameters through the unsupervised pre-training of deep belief networks (DBN), therefore improving the detection accuracy. It is demonstrated with experimental results that the proposed technique can provide a real-time response to the attack with a significantly improved detection ratio in controller area network (CAN) bus.

Intrusion Detection System Using Deep Neural Network for In-Vehicle Network Security

PubMed Central

Kang, Min-Joo

2016-01-01

A novel intrusion detection system (IDS) using a deep neural network (DNN) is proposed to enhance the security of in-vehicular network. The parameters building the DNN structure are trained with probability-based feature vectors that are extracted from the in-vehicular network packets. For a given packet, the DNN provides the probability of each class discriminating normal and attack packets, and, thus the sensor can identify any malicious attack to the vehicle. As compared to the traditional artificial neural network applied to the IDS, the proposed technique adopts recent advances in deep learning studies such as initializing the parameters through the unsupervised pre-training of deep belief networks (DBN), therefore improving the detection accuracy. It is demonstrated with experimental results that the proposed technique can provide a real-time response to the attack with a significantly improved detection ratio in controller area network (CAN) bus. PMID:27271802

The deep space network

NASA Technical Reports Server (NTRS)

1975-01-01

The objectives, functions, and organization of the Deep Space Network are summarized along with deep space station, ground communication, and network operations control capabilities. Mission support of ongoing planetary/interplanetary flight projects is discussed with emphasis on Viking orbiter radio frequency compatibility tests, the Pioneer Venus orbiter mission, and Helios-1 mission status and operations. Progress is also reported in tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations.

The Deep Space Network

NASA Technical Reports Server (NTRS)

1974-01-01

The objectives, functions, and organization, of the Deep Space Network are summarized. Deep Space stations, ground communications, and network operations control capabilities are described. The network is designed for two-way communications with unmanned spacecraft traveling approximately 1600 km from earth to the farthest planets in the solar system. It has provided tracking and data acquisition support for the following projects: Ranger, Surveyor, Mariner, Pioneer, Apollo, Helios, Viking, and the Lunar Orbiter.

DANoC: An Efficient Algorithm and Hardware Codesign of Deep Neural Networks on Chip.

PubMed

Zhou, Xichuan; Li, Shengli; Tang, Fang; Hu, Shengdong; Lin, Zhi; Zhang, Lei

2017-07-18

Deep neural networks (NNs) are the state-of-the-art models for understanding the content of images and videos. However, implementing deep NNs in embedded systems is a challenging task, e.g., a typical deep belief network could exhaust gigabytes of memory and result in bandwidth and computational bottlenecks. To address this challenge, this paper presents an algorithm and hardware codesign for efficient deep neural computation. A hardware-oriented deep learning algorithm, named the deep adaptive network, is proposed to explore the sparsity of neural connections. By adaptively removing the majority of neural connections and robustly representing the reserved connections using binary integers, the proposed algorithm could save up to 99.9% memory utility and computational resources without undermining classification accuracy. An efficient sparse-mapping-memory-based hardware architecture is proposed to fully take advantage of the algorithmic optimization. Different from traditional Von Neumann architecture, the deep-adaptive network on chip (DANoC) brings communication and computation in close proximity to avoid power-hungry parameter transfers between on-board memory and on-chip computational units. Experiments over different image classification benchmarks show that the DANoC system achieves competitively high accuracy and efficiency comparing with the state-of-the-art approaches.

DeepMitosis: Mitosis detection via deep detection, verification and segmentation networks.

PubMed

Li, Chao; Wang, Xinggang; Liu, Wenyu; Latecki, Longin Jan

2018-04-01

Mitotic count is a critical predictor of tumor aggressiveness in the breast cancer diagnosis. Nowadays mitosis counting is mainly performed by pathologists manually, which is extremely arduous and time-consuming. In this paper, we propose an accurate method for detecting the mitotic cells from histopathological slides using a novel multi-stage deep learning framework. Our method consists of a deep segmentation network for generating mitosis region when only a weak label is given (i.e., only the centroid pixel of mitosis is annotated), an elaborately designed deep detection network for localizing mitosis by using contextual region information, and a deep verification network for improving detection accuracy by removing false positives. We validate the proposed deep learning method on two widely used Mitosis Detection in Breast Cancer Histological Images (MITOSIS) datasets. Experimental results show that we can achieve the highest F-score on the MITOSIS dataset from ICPR 2012 grand challenge merely using the deep detection network. For the ICPR 2014 MITOSIS dataset that only provides the centroid location of mitosis, we employ the segmentation model to estimate the bounding box annotation for training the deep detection network. We also apply the verification model to eliminate some false positives produced from the detection model. By fusing scores of the detection and verification models, we achieve the state-of-the-art results. Moreover, our method is very fast with GPU computing, which makes it feasible for clinical practice. Copyright © 2018 Elsevier B.V. All rights reserved.

Development and Validation of a Deep Neural Network Model for Prediction of Postoperative In-hospital Mortality.

PubMed

Lee, Christine K; Hofer, Ira; Gabel, Eilon; Baldi, Pierre; Cannesson, Maxime

2018-04-17

The authors tested the hypothesis that deep neural networks trained on intraoperative features can predict postoperative in-hospital mortality. The data used to train and validate the algorithm consists of 59,985 patients with 87 features extracted at the end of surgery. Feed-forward networks with a logistic output were trained using stochastic gradient descent with momentum. The deep neural networks were trained on 80% of the data, with 20% reserved for testing. The authors assessed improvement of the deep neural network by adding American Society of Anesthesiologists (ASA) Physical Status Classification and robustness of the deep neural network to a reduced feature set. The networks were then compared to ASA Physical Status, logistic regression, and other published clinical scores including the Surgical Apgar, Preoperative Score to Predict Postoperative Mortality, Risk Quantification Index, and the Risk Stratification Index. In-hospital mortality in the training and test sets were 0.81% and 0.73%. The deep neural network with a reduced feature set and ASA Physical Status classification had the highest area under the receiver operating characteristics curve, 0.91 (95% CI, 0.88 to 0.93). The highest logistic regression area under the curve was found with a reduced feature set and ASA Physical Status (0.90, 95% CI, 0.87 to 0.93). The Risk Stratification Index had the highest area under the receiver operating characteristics curve, at 0.97 (95% CI, 0.94 to 0.99). Deep neural networks can predict in-hospital mortality based on automatically extractable intraoperative data, but are not (yet) superior to existing methods.

Diabetic retinopathy screening using deep neural network.

PubMed

Ramachandran, Nishanthan; Hong, Sheng Chiong; Sime, Mary J; Wilson, Graham A

2017-09-07

There is a burgeoning interest in the use of deep neural network in diabetic retinal screening. To determine whether a deep neural network could satisfactorily detect diabetic retinopathy that requires referral to an ophthalmologist from a local diabetic retinal screening programme and an international database. Retrospective audit. Diabetic retinal photos from Otago database photographed during October 2016 (485 photos), and 1200 photos from Messidor international database. Receiver operating characteristic curve to illustrate the ability of a deep neural network to identify referable diabetic retinopathy (moderate or worse diabetic retinopathy or exudates within one disc diameter of the fovea). Area under the receiver operating characteristic curve, sensitivity and specificity. For detecting referable diabetic retinopathy, the deep neural network had an area under receiver operating characteristic curve of 0.901 (95% confidence interval 0.807-0.995), with 84.6% sensitivity and 79.7% specificity for Otago and 0.980 (95% confidence interval 0.973-0.986), with 96.0% sensitivity and 90.0% specificity for Messidor. This study has shown that a deep neural network can detect referable diabetic retinopathy with sensitivities and specificities close to or better than 80% from both an international and a domestic (New Zealand) database. We believe that deep neural networks can be integrated into community screening once they can successfully detect both diabetic retinopathy and diabetic macular oedema. © 2017 Royal Australian and New Zealand College of Ophthalmologists.

The deep space network

NASA Technical Reports Server (NTRS)

1980-01-01

The functions and facilities of the Deep Space Network are considered. Progress in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations is reported.

The deep space network

NASA Technical Reports Server (NTRS)

1979-01-01

Progress is reported in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations. The functions and facilities of the Deep Space Network are emphasized.

Automatic recognition of holistic functional brain networks using iteratively optimized convolutional neural networks (IO-CNN) with weak label initialization.

PubMed

Zhao, Yu; Ge, Fangfei; Liu, Tianming

2018-07-01

fMRI data decomposition techniques have advanced significantly from shallow models such as Independent Component Analysis (ICA) and Sparse Coding and Dictionary Learning (SCDL) to deep learning models such Deep Belief Networks (DBN) and Convolutional Autoencoder (DCAE). However, interpretations of those decomposed networks are still open questions due to the lack of functional brain atlases, no correspondence across decomposed or reconstructed networks across different subjects, and significant individual variabilities. Recent studies showed that deep learning, especially deep convolutional neural networks (CNN), has extraordinary ability of accommodating spatial object patterns, e.g., our recent works using 3D CNN for fMRI-derived network classifications achieved high accuracy with a remarkable tolerance for mistakenly labelled training brain networks. However, the training data preparation is one of the biggest obstacles in these supervised deep learning models for functional brain network map recognitions, since manual labelling requires tedious and time-consuming labours which will sometimes even introduce label mistakes. Especially for mapping functional networks in large scale datasets such as hundreds of thousands of brain networks used in this paper, the manual labelling method will become almost infeasible. In response, in this work, we tackled both the network recognition and training data labelling tasks by proposing a new iteratively optimized deep learning CNN (IO-CNN) framework with an automatic weak label initialization, which enables the functional brain networks recognition task to a fully automatic large-scale classification procedure. Our extensive experiments based on ABIDE-II 1099 brains' fMRI data showed the great promise of our IO-CNN framework. Copyright © 2018 Elsevier B.V. All rights reserved.

Social networks of adults with an intellectual disability from South Asian and White communities in the United Kingdom: A comparison.

PubMed

Bhardwaj, Anjali K; Forrester-Jones, Rachel V E; Murphy, Glynis H

2018-03-01

Little research exists comparing the social networks of people with intellectual disability (ID) from South Asian and White backgrounds. This UK study reports on the barriers that South Asian people with intellectual disability face in relation to social inclusion compared to their White counterparts. A mixed-methods research design was adopted to explore the social lives of 27 men (15 White; 12 South Asian) and 20 women (10 White; 10 South Asian with intellectual disability). Descriptive and parametric tests were used to analyse the quantitative data. The average network size of the whole group was 32 members. South Asian participants had more family members whilst White participants had more service users and staff in their networks; 96% network members from White intellectual disability group were also of White background, whilst the South Asian group had mixed ethnic network members. Social networks of individuals with intellectual disability in this study were found to be larger overall in comparison with previous studies, whilst network structure differed between the White and South Asian population. These differences have implications relating to future service planning and appropriateness of available facilities. © 2017 John Wiley & Sons Ltd.

Considering the Role of Stress in Populations of High-Risk, Underserved Community Networks Program Centers

PubMed Central

Hébert, James R.; Braun, Kathryn L.; Kaholokula, Joseph Keawe‘aimoku; Armstead, Cheryl A.; Burch, James B.; Thompson, Beti

2015-01-01

Background Cancer disparities are associated with a broad range of sociocultural determinants of health that operate in community contexts. High-risk populations may be more vulnerable to social and environmental factors that lead to chronic stress. Theoretical and empirical research indicates that exposure to contextual and sociocultural stress alters biological systems, thereby influencing cancer risk, progression, and, ultimately, mortality. Objective We sought to describe contextual pathways through which stress likely increases cancer risk in high-risk, underserved populations. Methods This review presents a description of the link between contextual stressors and disease risk disparities within underserved communities, with a focus on 1) stress as a proximal link between biological processes, such as cytokine responses, inflammation, and cancer and 2) stress as a distal link to cancer through biobehavioral risk factors such as poor diet, physical inactivity, circadian rhythm or sleep disruption, and substance abuse. These concepts are illustrated through application to populations served by three National Cancer Institute-funded Community Networks Program Centers (CNPCs): African Americans in the Deep South (the South Carolina Cancer Disparities Community Network [SCCDCN]), Native Hawaiians (‘Imi Hale—Native Hawaiian Cancer Network), and Latinos in the Lower Yakima Valley of Washington State (The Center for Hispanic Health Promotion: Reducing Cancer Disparities). Conclusions Stress experienced by the underserved communities represented in the CNPCs is marked by social, biological, and behavioral pathways that increase cancer risk. A case is presented to increase research on sociocultural determinants of health, stress, and cancer risk among racial/ethnic minorities in underserved communities. PMID:26213406

Regional hydrology and simulation of deep ground-water flow in the Southeastern Coastal Plain aquifer system in Mississippi, Alabama, Georgia, and South Carolina

USGS Publications Warehouse

Barker, R.A.; Pernik, Maribeth

1994-01-01

The Southeastern Coastal Plain aquifer system is a coastward-sloping, wedge-shaped sand and gravel reservoir exposed in outcrop to a humid climate and drained by an extensive surface-water network. Ground-water pumpage has increased to about 765 cubic feet per second since 1900, causing water-level declines of more than 150 feet in places, while base flow to major streams has decreased about 350 cubic feet per second. The water-level declines and adjustments in recharge and discharge are not expected to seriously restrict future ground-water development.

The deep space network

NASA Technical Reports Server (NTRS)

1979-01-01

A report is given of the Deep Space Networks progress in (1) flight project support, (2) tracking and data acquisition research and technology, (3) network engineering, (4) hardware and software implementation, and (5) operations.

Magnetoencephalographic imaging of deep corticostriatal network activity during a rewards paradigm.

PubMed

Kanal, Eliezer Y; Sun, Mingui; Ozkurt, Tolga E; Jia, Wenyan; Sclabassi, Robert

2009-01-01

The human rewards network is a complex system spanning both cortical and subcortical regions. While much is known about the functions of the various components of the network, research on the behavior of the network as a whole has been stymied due to an inability to detect signals at a high enough temporal resolution from both superficial and deep network components simultaneously. In this paper, we describe the application of magnetoencephalographic imaging (MEG) combined with advanced signal processing techniques to this problem. Using data collected while subjects performed a rewards-related gambling paradigm demonstrated to activate the rewards network, we were able to identify neural signals which correspond to deep network activity. We also show that this signal was not observable prior to filtration. These results suggest that MEG imaging may be a viable tool for the detection of deep neural activity.

Robust hepatic vessel segmentation using multi deep convolution network

NASA Astrophysics Data System (ADS)

Kitrungrotsakul, Titinunt; Han, Xian-Hua; Iwamoto, Yutaro; Foruzan, Amir Hossein; Lin, Lanfen; Chen, Yen-Wei

2017-03-01

Extraction of blood vessels of the organ is a challenging task in the area of medical image processing. It is really difficult to get accurate vessel segmentation results even with manually labeling by human being. The difficulty of vessels segmentation is the complicated structure of blood vessels and its large variations that make them hard to recognize. In this paper, we present deep artificial neural network architecture to automatically segment the hepatic vessels from computed tomography (CT) image. We proposed novel deep neural network (DNN) architecture for vessel segmentation from a medical CT volume, which consists of three deep convolution neural networks to extract features from difference planes of CT data. The three networks have share features at the first convolution layer but will separately learn their own features in the second layer. All three networks will join again at the top layer. To validate effectiveness and efficiency of our proposed method, we conduct experiments on 12 CT volumes which training data are randomly generate from 5 CT volumes and 7 using for test. Our network can yield an average dice coefficient 0.830, while 3D deep convolution neural network can yield around 0.7 and multi-scale can yield only 0.6.

Future Plans for NASA's Deep Space Network

NASA Technical Reports Server (NTRS)

Deutsch, Leslie J.; Preston, Robert A.; Geldzahler, Barry J.

2008-01-01

This slide presentation reviews the importance of NASA's Deep Space Network (DSN) to space exploration, and future planned improvements to the communication capabilities that the network allows, in terms of precision, and communication power.

The deep space network

NASA Technical Reports Server (NTRS)

1977-01-01

The facilities, programming system, and monitor and control system for the deep space network are described. Ongoing planetary and interplanetary flight projects are reviewed, along with tracking and ground-based navigation, communications, and network and facility engineering.

The diurnal interaction between convection and peninsular-scale forcing over South Florida

NASA Technical Reports Server (NTRS)

Cooper, H. J.; Simpson, J.; Garstang, M.

1982-01-01

One of the outstanding problems in modern meterology is that of describing in detail the manner in which larger scales of motion interact with, influence and are influenced by successively smaller scales of motion. The present investigation is concerned with a study of the diurnal evolution of convection, the interaction between the peninsular-scale convergence and convection, and the role of the feedback produced by the cloud-scale downdrafts in the maintenance of the convection. Attention is given to the analysis, the diurnal cycle of the network area-averaged divergence, convective-scale divergence, convective mass transports, and the peninsular scale divergence. The links established in the investigation between the large scale (peninsular), the mesoscale (network), and the convective scale (cloud) are found to be of fundamental importance to the understanding of the initiation, maintenance, and decay of deep precipitating convection and to its theoretical parameterization.

Office of Tracking and Data Acquisition. [deep space network and spacecraft tracking

NASA Technical Reports Server (NTRS)

1975-01-01

The Office of Tracking and Data Acquisition (OTDA) and its two worldwide tracking network facilities, the Spaceflight Tracking and Data Network and the Deep Space Network, are described. Other topics discussed include the NASA communications network, the tracking and data relay satellite system, other OTDA tracking activities, and OTDA milestones.

The Future of the Deep Space Network: Technology Development for K2-Band Deep Space Communications

NASA Technical Reports Server (NTRS)

Bhanji, Alaudin M.

1999-01-01

Projections indicate that in the future the number of NASA's robotic deep space missions is likely to increase significantly. A launch rate of up to 4-6 launches per year is projected with up to 25 simultaneous missions active [I]. Future high resolution mapping missions to other planetary bodies as well as other experiments are likely to require increased downlink capacity. These future deep space communications requirements will, according to baseline loading analysis, exceed the capacity of NASA's Deep Space Network in its present form. There are essentially two approaches for increasing the channel capacity of the Deep Space Network. Given the near-optimum performance of the network at the two deep space communications bands, S-Band (uplink 2.025-2.120 GHz, downlink 2.2-2.3 GHz), and X-Band (uplink 7.145-7.19 GHz, downlink 8.48.5 GHz), additional improvements bring only marginal return for the investment. Thus the only way to increase channel capacity is simply to construct more antennas, receivers, transmitters and other hardware. This approach is relatively low-risk but involves increasing both the number of assets in the network and operational costs.

Plant Species Identification by Bi-channel Deep Convolutional Networks

NASA Astrophysics Data System (ADS)

He, Guiqing; Xia, Zhaoqiang; Zhang, Qiqi; Zhang, Haixi; Fan, Jianping

2018-04-01

Plant species identification achieves much attention recently as it has potential application in the environmental protection and human life. Although deep learning techniques can be directly applied for plant species identification, it still needs to be designed for this specific task to obtain the state-of-art performance. In this paper, a bi-channel deep learning framework is developed for identifying plant species. In the framework, two different sub-networks are fine-tuned over their pretrained models respectively. And then a stacking layer is used to fuse the output of two different sub-networks. We construct a plant dataset of Orchidaceae family for algorithm evaluation. Our experimental results have demonstrated that our bi-channel deep network can achieve very competitive performance on accuracy rates compared to the existing deep learning algorithm.

DeepQA: improving the estimation of single protein model quality with deep belief networks.

PubMed

Cao, Renzhi; Bhattacharya, Debswapna; Hou, Jie; Cheng, Jianlin

2016-12-05

Protein quality assessment (QA) useful for ranking and selecting protein models has long been viewed as one of the major challenges for protein tertiary structure prediction. Especially, estimating the quality of a single protein model, which is important for selecting a few good models out of a large model pool consisting of mostly low-quality models, is still a largely unsolved problem. We introduce a novel single-model quality assessment method DeepQA based on deep belief network that utilizes a number of selected features describing the quality of a model from different perspectives, such as energy, physio-chemical characteristics, and structural information. The deep belief network is trained on several large datasets consisting of models from the Critical Assessment of Protein Structure Prediction (CASP) experiments, several publicly available datasets, and models generated by our in-house ab initio method. Our experiments demonstrate that deep belief network has better performance compared to Support Vector Machines and Neural Networks on the protein model quality assessment problem, and our method DeepQA achieves the state-of-the-art performance on CASP11 dataset. It also outperformed two well-established methods in selecting good outlier models from a large set of models of mostly low quality generated by ab initio modeling methods. DeepQA is a useful deep learning tool for protein single model quality assessment and protein structure prediction. The source code, executable, document and training/test datasets of DeepQA for Linux is freely available to non-commercial users at http://cactus.rnet.missouri.edu/DeepQA/ .

The Deep Space Network: A Radio Communications Instrument for Deep Space Exploration

NASA Technical Reports Server (NTRS)

Renzetti, N. A.; Stelzried, C. T.; Noreen, G. K.; Slobin, S. D.; Petty, S. M.; Trowbridge, D. L.; Donnelly, H.; Kinman, P. W.; Armstrong, J. W.; Burow, N. A.

1983-01-01

The primary purpose of the Deep Space Network (DSN) is to serve as a communications instrument for deep space exploration, providing communications between the spacecraft and the ground facilities. The uplink communications channel provides instructions or commands to the spacecraft. The downlink communications channel provides command verification and spacecraft engineering and science instrument payload data.

Personal social networks and organizational affiliation of South Asians in the United States.

PubMed

Kandula, Namratha R; Cooper, Andrew J; Schneider, John A; Fujimoto, Kayo; Kanaya, Alka M; Van Horn, Linda; deKoning, Lawrence; Siddique, Juned

2018-02-05

Understanding the social lives of South Asian immigrants in the United States (U.S) and their influence on health can inform interpersonal and community-level health interventions for this growing community. This paper describe the rationale, survey design, measurement, and network properties of 700 South Asian individuals in the Mediators of Atherosclerosis in South Asians Living in America (MASALA) social networks ancillary study. MASALA is a community-based cohort, established in 2010, to understand risk factors for cardiovascular disease among South Asians living in the U.S. Survey data collection on personal social networks occurred between 2014 and 2017. Network measurements included size, composition, density, and organizational affiliations. Data on participants' self-rated health and social support functions and health-related discussions among network members were also collected. Participants' age ranged from 44 to 84 (average 59 years), and 57% were men. South Asians had large (size=5.6, SD=2.6), kin-centered (proportion kin=0.71, SD=0.28), and dense networks. Affiliation with religious and spiritual organizations was perceived as beneficial to health. Emotional closeness with network members was positively associated with participants' self-rated health (p-value <0.001), and networks with higher density and more kin were significantly associated with health-related discussions. The MASALA networks study advances research on the cultural patterning of social relationships and sources of social support in South Asians living in the U.S. Future analyses will examine how personal social networks and organizational affiliations influence South Asians' health behaviors and outcomes. ClinicalTrials.gov identifier: NCT02268513.

The deep space network, Volume 11

NASA Technical Reports Server (NTRS)

1972-01-01

Deep Space Network progress in flight project support, Tracking and Data Acquisition research and technology, network engineering, hardware and software implementation, and operations are presented. Material is presented in each of the following categories: description of DSN; mission support; radio science; support research and technology; network engineering and implementation; and operations and facilities.

Breakup of last glacial deep stratification in the South Pacific

NASA Astrophysics Data System (ADS)

Basak, Chandranath; Fröllje, Henning; Lamy, Frank; Gersonde, Rainer; Benz, Verena; Anderson, Robert F.; Molina-Kescher, Mario; Pahnke, Katharina

2018-02-01

Stratification of the deep Southern Ocean during the Last Glacial Maximum is thought to have facilitated carbon storage and subsequent release during the deglaciation as stratification broke down, contributing to atmospheric CO2 rise. Here, we present neodymium isotope evidence from deep to abyssal waters in the South Pacific that confirms stratification of the deepwater column during the Last Glacial Maximum. The results indicate a glacial northward expansion of Ross Sea Bottom Water and a Southern Hemisphere climate trigger for the deglacial breakup of deep stratification. It highlights the important role of abyssal waters in sustaining a deep glacial carbon reservoir and Southern Hemisphere climate change as a prerequisite for the destabilization of the water column and hence the deglacial release of sequestered CO2 through upwelling.

An improved advertising CTR prediction approach based on the fuzzy deep neural network

PubMed Central

Gao, Shu; Li, Mingjiang

2018-01-01

Combining a deep neural network with fuzzy theory, this paper proposes an advertising click-through rate (CTR) prediction approach based on a fuzzy deep neural network (FDNN). In this approach, fuzzy Gaussian-Bernoulli restricted Boltzmann machine (FGBRBM) is first applied to input raw data from advertising datasets. Next, fuzzy restricted Boltzmann machine (FRBM) is used to construct the fuzzy deep belief network (FDBN) with the unsupervised method layer by layer. Finally, fuzzy logistic regression (FLR) is utilized for modeling the CTR. The experimental results show that the proposed FDNN model outperforms several baseline models in terms of both data representation capability and robustness in advertising click log datasets with noise. PMID:29727443

Deep multi-scale convolutional neural network for hyperspectral image classification

NASA Astrophysics Data System (ADS)

Zhang, Feng-zhe; Yang, Xia

2018-04-01

In this paper, we proposed a multi-scale convolutional neural network for hyperspectral image classification task. Firstly, compared with conventional convolution, we utilize multi-scale convolutions, which possess larger respective fields, to extract spectral features of hyperspectral image. We design a deep neural network with a multi-scale convolution layer which contains 3 different convolution kernel sizes. Secondly, to avoid overfitting of deep neural network, dropout is utilized, which randomly sleeps neurons, contributing to improve the classification accuracy a bit. In addition, new skills like ReLU in deep learning is utilized in this paper. We conduct experiments on University of Pavia and Salinas datasets, and obtained better classification accuracy compared with other methods.

An improved advertising CTR prediction approach based on the fuzzy deep neural network.

PubMed

Jiang, Zilong; Gao, Shu; Li, Mingjiang

2018-01-01

Combining a deep neural network with fuzzy theory, this paper proposes an advertising click-through rate (CTR) prediction approach based on a fuzzy deep neural network (FDNN). In this approach, fuzzy Gaussian-Bernoulli restricted Boltzmann machine (FGBRBM) is first applied to input raw data from advertising datasets. Next, fuzzy restricted Boltzmann machine (FRBM) is used to construct the fuzzy deep belief network (FDBN) with the unsupervised method layer by layer. Finally, fuzzy logistic regression (FLR) is utilized for modeling the CTR. The experimental results show that the proposed FDNN model outperforms several baseline models in terms of both data representation capability and robustness in advertising click log datasets with noise.

A Study of Highly Qualified Science Teachers' Career Trajectory in the Deep, Rural South: Examining a Link between Deprofessionalization and Teacher Dissatisfaction

ERIC Educational Resources Information Center

Hodges, Georgia W.; Tippins, Deborah; Oliver, J. Steve

2013-01-01

Science teacher retention, attrition, and migration continue to perplex educational scholars, political entities, as well as the general public. This study utilized an interpretive methodological design to generate assertions regarding career choice made by highly qualified science teachers in the deep, rural South through analysis of documents,…

Spatiotemporal Recurrent Convolutional Networks for Traffic Prediction in Transportation Networks

PubMed Central

Yu, Haiyang; Wu, Zhihai; Wang, Shuqin; Wang, Yunpeng; Ma, Xiaolei

2017-01-01

Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs), for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs) and long short-term memory (LSTM) neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction. PMID:28672867

Spatiotemporal Recurrent Convolutional Networks for Traffic Prediction in Transportation Networks.

PubMed

Yu, Haiyang; Wu, Zhihai; Wang, Shuqin; Wang, Yunpeng; Ma, Xiaolei

2017-06-26

Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs), for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs) and long short-term memory (LSTM) neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction.

Upstream ecological risks for overweight and obesity among African American youth in a rural town in the Deep South, 2007.

PubMed

Scott, Alison J; Wilson, Rebecca F

2011-01-01

Few studies have focused on overweight and obesity among rural African American youth in the Deep South, despite disproportionately high rates in this group. In addition, few studies have been conducted to elucidate how these disparities are created and perpetuated within rural communities in this region. This descriptive study explores community-based risks for overweight and obesity among African American youth in a rural town in the Deep South. We used ecological theory in conjunction with embodiment theory to explore how upstream ecological factors may contribute to risk of overweight and obesity for African American youth in a rural town in the Deep South. We conducted and analyzed in-depth interviews with African American community members who interact with youth in varying contexts (home, school, church, community). Participants most commonly stated that race relations, poverty, and the built environment were barriers to maintaining a healthy weight. Findings suggested the need for rural, community-based interventions that target obesity at multiple ecological levels and incorporate issues related to race, poverty, and the built environment. More research is needed to determine how disparities in obesity are created and perpetuated in specific community contexts.

Salient object detection based on multi-scale contrast.

PubMed

Wang, Hai; Dai, Lei; Cai, Yingfeng; Sun, Xiaoqiang; Chen, Long

2018-05-01

Due to the development of deep learning networks, a salient object detection based on deep learning networks, which are used to extract the features, has made a great breakthrough compared to the traditional methods. At present, the salient object detection mainly relies on very deep convolutional network, which is used to extract the features. In deep learning networks, an dramatic increase of network depth may cause more training errors instead. In this paper, we use the residual network to increase network depth and to mitigate the errors caused by depth increase simultaneously. Inspired by image simplification, we use color and texture features to obtain simplified image with multiple scales by means of region assimilation on the basis of super-pixels in order to reduce the complexity of images and to improve the accuracy of salient target detection. We refine the feature on pixel level by the multi-scale feature correction method to avoid the feature error when the image is simplified at the above-mentioned region level. The final full connection layer not only integrates features of multi-scale and multi-level but also works as classifier of salient targets. The experimental results show that proposed model achieves better results than other salient object detection models based on original deep learning networks. Copyright © 2018 Elsevier Ltd. All rights reserved.

Student beats the teacher: deep neural networks for lateral ventricles segmentation in brain MR

NASA Astrophysics Data System (ADS)

Ghafoorian, Mohsen; Teuwen, Jonas; Manniesing, Rashindra; Leeuw, Frank-Erik d.; van Ginneken, Bram; Karssemeijer, Nico; Platel, Bram

2018-03-01

Ventricular volume and its progression are known to be linked to several brain diseases such as dementia and schizophrenia. Therefore accurate measurement of ventricle volume is vital for longitudinal studies on these disorders, making automated ventricle segmentation algorithms desirable. In the past few years, deep neural networks have shown to outperform the classical models in many imaging domains. However, the success of deep networks is dependent on manually labeled data sets, which are expensive to acquire especially for higher dimensional data in the medical domain. In this work, we show that deep neural networks can be trained on muchcheaper-to-acquire pseudo-labels (e.g., generated by other automated less accurate methods) and still produce more accurate segmentations compared to the quality of the labels. To show this, we use noisy segmentation labels generated by a conventional region growing algorithm to train a deep network for lateral ventricle segmentation. Then on a large manually annotated test set, we show that the network significantly outperforms the conventional region growing algorithm which was used to produce the training labels for the network. Our experiments report a Dice Similarity Coefficient (DSC) of 0.874 for the trained network compared to 0.754 for the conventional region growing algorithm (p < 0.001).

Diverse deep-sea fungi from the South China Sea and their antimicrobial activity.

PubMed

Zhang, Xiao-Yong; Zhang, Yun; Xu, Xin-Ya; Qi, Shu-Hua

2013-11-01

We investigated the diversity of fungal communities in nine different deep-sea sediment samples of the South China Sea by culture-dependent methods followed by analysis of fungal internal transcribed spacer (ITS) sequences. Although 14 out of 27 identified species were reported in a previous study, 13 species were isolated from sediments of deep-sea environments for the first report. Moreover, these ITS sequences of six isolates shared 84-92 % similarity with their closest matches in GenBank, which suggested that they might be novel phylotypes of genera Ajellomyces, Podosordaria, Torula, and Xylaria. The antimicrobial activities of these fungal isolates were explored using a double-layer technique. A relatively high proportion (56 %) of fungal isolates exhibited antimicrobial activity against at least one pathogenic bacterium or fungus among four marine pathogenic microbes (Micrococcus luteus, Pseudoaltermonas piscida, Aspergerillus versicolor, and A. sydowii). Out of these antimicrobial fungi, the genera Arthrinium, Aspergillus, and Penicillium exhibited antibacterial and antifungal activities, while genus Aureobasidium displayed only antibacterial activity, and genera Acremonium, Cladosporium, Geomyces, and Phaeosphaeriopsis displayed only antifungal activity. To our knowledge, this is the first report to investigate the diversity and antimicrobial activity of culturable deep-sea-derived fungi in the South China Sea. These results suggest that diverse deep-sea fungi from the South China Sea are a potential source for antibiotics' discovery and further increase the pool of fungi available for natural bioactive product screening.

White blood cells identification system based on convolutional deep neural learning networks.

PubMed

Shahin, A I; Guo, Yanhui; Amin, K M; Sharawi, Amr A

2017-11-16

White blood cells (WBCs) differential counting yields valued information about human health and disease. The current developed automated cell morphology equipments perform differential count which is based on blood smear image analysis. Previous identification systems for WBCs consist of successive dependent stages; pre-processing, segmentation, feature extraction, feature selection, and classification. There is a real need to employ deep learning methodologies so that the performance of previous WBCs identification systems can be increased. Classifying small limited datasets through deep learning systems is a major challenge and should be investigated. In this paper, we propose a novel identification system for WBCs based on deep convolutional neural networks. Two methodologies based on transfer learning are followed: transfer learning based on deep activation features and fine-tuning of existed deep networks. Deep acrivation featues are extracted from several pre-trained networks and employed in a traditional identification system. Moreover, a novel end-to-end convolutional deep architecture called "WBCsNet" is proposed and built from scratch. Finally, a limited balanced WBCs dataset classification is performed through the WBCsNet as a pre-trained network. During our experiments, three different public WBCs datasets (2551 images) have been used which contain 5 healthy WBCs types. The overall system accuracy achieved by the proposed WBCsNet is (96.1%) which is more than different transfer learning approaches or even the previous traditional identification system. We also present features visualization for the WBCsNet activation which reflects higher response than the pre-trained activated one. a novel WBCs identification system based on deep learning theory is proposed and a high performance WBCsNet can be employed as a pre-trained network. Copyright © 2017. Published by Elsevier B.V.

A three-dimensional geological reconstruction of Noctis Labyrinthus slope tectonics from CaSSIS data

NASA Astrophysics Data System (ADS)

Massironi, M. M.; Pozzobon, R. P.; Lucchetti, A. L.; Simioni, E. S.; Re, C. R.; Mudrič, T. M.; Pajola, M. P.; Cremonese, G. C.; Pommerol, A. P.; Salese, F. S.; Thomas, N. T.; Mege, D. M.

2017-09-01

In November 2016 the CaSSIS (Colour and Stereo Surface Imaging System) imaging system onboard the European Space Agency's ExoMars Trace Gas Orbiter (TGO) acquired 18 images (each composed by 30 framelets for each of the 4 colour channels) of the Martian surface. The first stereo- pairs were taken during the closest approach, at a distance of 520 km from the surface, over the Hebes Chasma and Noctis Labyrithus regions. In the latter case a DTM was prepared over a north facing slope bounding to the north a 2000 m deep depression and to the south a plateau complicated by extensional fault networks. Such slope is characterised by a downthrown block that can be interpreted as a Deep Seated Gravitational Slope Deformation (DSGSD) sensu. In this work we will present a 3D geological reconstruction of the phenomenon that allowed us to constrain the possible main sliding surface, the volumes involved in the gravitational process and the kinematics of the mass movement.

Deep learning for predicting the monsoon over the homogeneous regions of India

NASA Astrophysics Data System (ADS)

Saha, Moumita; Mitra, Pabitra; Nanjundiah, Ravi S.

2017-06-01

Indian monsoon varies in its nature over the geographical regions. Predicting the rainfall not just at the national level, but at the regional level is an important task. In this article, we used a deep neural network, namely, the stacked autoencoder to automatically identify climatic factors that are capable of predicting the rainfall over the homogeneous regions of India. An ensemble regression tree model is used for monsoon prediction using the identified climatic predictors. The proposed model provides forecast of the monsoon at a long lead time which supports the government to implement appropriate policies for the economic growth of the country. The monsoon of the central, north-east, north-west, and south-peninsular India regions are predicted with errors of 4.1%, 5.1%, 5.5%, and 6.4%, respectively. The identified predictors show high skill in predicting the regional monsoon having high variability. The proposed model is observed to be competitive with the state-of-the-art prediction models.

Deep learning in the small sample size setting: cascaded feed forward neural networks for medical image segmentation

NASA Astrophysics Data System (ADS)

Gaonkar, Bilwaj; Hovda, David; Martin, Neil; Macyszyn, Luke

2016-03-01

Deep Learning, refers to large set of neural network based algorithms, have emerged as promising machine- learning tools in the general imaging and computer vision domains. Convolutional neural networks (CNNs), a specific class of deep learning algorithms, have been extremely effective in object recognition and localization in natural images. A characteristic feature of CNNs, is the use of a locally connected multi layer topology that is inspired by the animal visual cortex (the most powerful vision system in existence). While CNNs, perform admirably in object identification and localization tasks, typically require training on extremely large datasets. Unfortunately, in medical image analysis, large datasets are either unavailable or are extremely expensive to obtain. Further, the primary tasks in medical imaging are organ identification and segmentation from 3D scans, which are different from the standard computer vision tasks of object recognition. Thus, in order to translate the advantages of deep learning to medical image analysis, there is a need to develop deep network topologies and training methodologies, that are geared towards medical imaging related tasks and can work in a setting where dataset sizes are relatively small. In this paper, we present a technique for stacked supervised training of deep feed forward neural networks for segmenting organs from medical scans. Each `neural network layer' in the stack is trained to identify a sub region of the original image, that contains the organ of interest. By layering several such stacks together a very deep neural network is constructed. Such a network can be used to identify extremely small regions of interest in extremely large images, inspite of a lack of clear contrast in the signal or easily identifiable shape characteristics. What is even more intriguing is that the network stack achieves accurate segmentation even when it is trained on a single image with manually labelled ground truth. We validate this approach,using a publicly available head and neck CT dataset. We also show that a deep neural network of similar depth, if trained directly using backpropagation, cannot acheive the tasks achieved using our layer wise training paradigm.

Saliency U-Net: A regional saliency map-driven hybrid deep learning network for anomaly segmentation

NASA Astrophysics Data System (ADS)

Karargyros, Alex; Syeda-Mahmood, Tanveer

2018-02-01

Deep learning networks are gaining popularity in many medical image analysis tasks due to their generalized ability to automatically extract relevant features from raw images. However, this can make the learning problem unnecessarily harder requiring network architectures of high complexity. In case of anomaly detection, in particular, there is often sufficient regional difference between the anomaly and the surrounding parenchyma that could be easily highlighted through bottom-up saliency operators. In this paper we propose a new hybrid deep learning network using a combination of raw image and such regional maps to more accurately learn the anomalies using simpler network architectures. Specifically, we modify a deep learning network called U-Net using both the raw and pre-segmented images as input to produce joint encoding (contraction) and expansion paths (decoding) in the U-Net. We present results of successfully delineating subdural and epidural hematomas in brain CT imaging and liver hemangioma in abdominal CT images using such network.

Statistical porcess control in Deep Space Network operation

NASA Technical Reports Server (NTRS)

Hodder, J. A.

2002-01-01

This report describes how the Deep Space Mission System (DSMS) Operations Program Office at the Jet Propulsion Laboratory's (EL) uses Statistical Process Control (SPC) to monitor performance and evaluate initiatives for improving processes on the National Aeronautics and Space Administration's (NASA) Deep Space Network (DSN).

Climatic records of the last and penultimate deglaciations in the South Atlantic and South Indian Ocean

NASA Astrophysics Data System (ADS)

Michel, Elisabeth; Waelbroeck, Claire; Govin, Aline; Skinner, Luke; Vàzquez Riveiros, Natalia; Dewilde, Fabien; Isguder, Gulay; Rebaubier, Hélène

2013-04-01

Surface and deep-water records of Termination I and II in two twin South Atlantic deep-sea cores (44°09' S, 14°14' W, 3770 m depth) and one South Indian core (46°29' S, 88°01' E, 3420 m depth) are presented. Sea surface temperature has been reconstructed based on planktonic foraminifera census counts in all cases, as well as Mg/Ca of G. bulloides and N. pachyderma s. over the last deglaciation. The uncertainty on reconstructed SST using different statistical methods and different faunal databases is assessed. Over the last deglaciation, combined 14C dating and correlation of the SST record with the air temperature signal recorded in Antarctic ice cores allowed us to correct for variable surface reservoir ages in the South Atlantic core (Skinner et al., 2010). Preliminary dating of the South Indian core over the last termination has been done by correlation of its magnetic signal with those of a neighboring 14C dated core (Smart et al., 2010). We have refined the later age scale using the Atlantic core age scale as reference. Benthic isotopic signals in the South Atlantic and South Indian cores over the last deglaciation exhibit the same amplitude and timing. Our results thus indicate that bottom waters at the South Indian site remained isolated from better ventilated deep waters of northern origin until ~15 ka (Waelbroeck et al., 2011). Over Termination II, the two cores have been dated by correlation of their SST records with the air temperature signal recorded in EDC versus the EDC3 age scale (Govin et al., 2009; 2012). A careful examination of the various sources of uncertainty on the derived dating has been performed. Benthic and planktonic isotopic signals reveal analogies but also differences with respect to the last termination. SST was significantly warmer during the Last Interglacial than during the Holocene in both sites. South Atlantic deep waters were also significantly better ventilated during the Last Interglacial than during the Holocene, whereas bottom water ventilation was similar during these two interglacials at the South Indian site.

Deep Constrained Siamese Hash Coding Network and Load-Balanced Locality-Sensitive Hashing for Near Duplicate Image Detection.

PubMed

Hu, Weiming; Fan, Yabo; Xing, Junliang; Sun, Liang; Cai, Zhaoquan; Maybank, Stephen

2018-09-01

We construct a new efficient near duplicate image detection method using a hierarchical hash code learning neural network and load-balanced locality-sensitive hashing (LSH) indexing. We propose a deep constrained siamese hash coding neural network combined with deep feature learning. Our neural network is able to extract effective features for near duplicate image detection. The extracted features are used to construct a LSH-based index. We propose a load-balanced LSH method to produce load-balanced buckets in the hashing process. The load-balanced LSH significantly reduces the query time. Based on the proposed load-balanced LSH, we design an effective and feasible algorithm for near duplicate image detection. Extensive experiments on three benchmark data sets demonstrate the effectiveness of our deep siamese hash encoding network and load-balanced LSH.

11. VIEW OF SOUTHWEST CORNER OF SOUTH WING OF TECHWOOD ...

Library of Congress Historic Buildings Survey, Historic Engineering Record, Historic Landscapes Survey

11. VIEW OF SOUTHWEST CORNER OF SOUTH WING OF TECHWOOD DORMITORY. WEST FRONT OF SOUTH WING OBSCURED BY DEEP SHADE. - Techwood Homes, McDaniel Dormitory, 581-587 Techwood Drive, Atlanta, Fulton County, GA

The TOPOMOD-ITN project: unravel the origin of Earth's topography from modelling deep-surface processes

NASA Astrophysics Data System (ADS)

Faccenna, C.; Funiciello, F.

2012-04-01

EC-Marie Curie Initial Training Networks (ITN) projects aim to improve the career perspectives of young generations of researchers. Institutions from both academic and industry sectors form a collaborative network to recruit research fellows and provide them with opportunities to undertake research in the context of a joint research training program. In this frame, TOPOMOD - one of the training activities of EPOS, the new-born European Research Infrastructure for Geosciences - is a funded ITN project designed to investigate and model how surface processes interact with crustal tectonics and mantle convection to originate and develop topography of the continents over a wide range of spatial and temporal scales. The multi-disciplinary approach combines geophysics, geochemistry, tectonics and structural geology with advanced geodynamic numerical/analog modelling. TOPOMOD involves 8 European research teams internationally recognized for their excellence in complementary fields of Earth Sciences (Roma TRE, Utrecht, GFZ, ETH, Cambridge, Durham, Rennes, Barcelona), to which are associated 5 research institutions (CNR-Italy, Univ. Parma, Univ. Lausanne, Univ. Montpellier, Univ. Mainz) , 3 high-technology enterprises (Malvern Instruments, TNO, G.O. Logical Consulting) and 1 large multinational oil and gas company (ENI). This unique network places emphasis in experience-based training increasing the impact and international visibility of European research in modeling. Long-term collaboration and synergy are established among the overmentioned research teams through 15 cross-disciplinary research projects that combine case studies in well-chosen target areas from the Mediterranean, the Middle and Far East, west Africa, and South America, with new developments in structural geology, geomorphology, seismology, geochemistry, InSAR, laboratory and numerical modelling of geological processes from the deep mantle to the surface. These multidisciplinary projects altogether aim to answer a key question in earth Sciences: how do deep and surface processes interact to shape and control the topographic evolution of our planet.

The BOrborema Deep Electromagnetic and Seismic (BODES) Experiment

NASA Astrophysics Data System (ADS)

Julià, J.; Garcia, X.; Medeiros, W. E.; Farias do Nascimento, A.

2015-12-01

The Borborema Province of NE Brazil is a large Precambrian domain of the Brazilian shield located in the Northeasternmost corner of South America. It is bounded by the Parnaíba basin to the West and by the São Francisco craton to the South. Its structuration in the Precambrian has been related to compressional processes during the Brasiliano-Pan African orogeny (600-550 Ma). In the Mesozoic, extensional stresses eventually leading to continental breakup, left a number of aborted rift basins within the Province. After continental breakup, the evolution of the Province was marked by episodes of uplift, which might have been coeval with episodes of Cenozoic volcanism. The most prominent expression of those uplift processes is the Borborema Plateau, an elliptically shaped topographic feature in the eastern half of the Province with maximum elevations of ~1200 m. The origin of uplift in the Plateau has been the focus of a number of multi-institutional and multi-disciplinary studies in the past few years, which have imaged the deep structure of the eastern Province with unprecedented detail. The origin of uplift in the western Province, which includes a superb example of basin inversion demonstrated by the ~1000 km elevations of the Chapada do Araripe, however, has been seldom investigated. With the goal of investigating the deep structure of the western Province, a temporary network of 10 collocated seismic and magnetotelluric stations was deployed in the region. The collocated stations were arranged in an approximately NS direction, with an interspation spacing of ~70 km and spanning a total length of ~600 km. The seismic stations consisted of broadband sensors (RefTek 151B-120 "Observer") sampling at 100 Hz and were deployed in January 2015; the MT stations consisted of long-period magnetotelluric (LEMI) systems, sampling at 1 Hz and 4 Hz, and were deployed in April 2015 for a period of ~2 weeks. Preliminary results based on teleseismic P-wave receiver functions suggest that the crust thickens towards the South, from 33 km in the Ceará domain to 44 km in the São Francisco craton. Preliminary analyis of MT data suggests a heterogeneous lithosphere, with marked changes in electrical properties around the Chapada do Araripe and a marked resistive structure towards the South, where the profile enters the São Francisco craton.

Detecting atrial fibrillation by deep convolutional neural networks.

PubMed

Xia, Yong; Wulan, Naren; Wang, Kuanquan; Zhang, Henggui

2018-02-01

Atrial fibrillation (AF) is the most common cardiac arrhythmia. The incidence of AF increases with age, causing high risks of stroke and increased morbidity and mortality. Efficient and accurate diagnosis of AF based on the ECG is valuable in clinical settings and remains challenging. In this paper, we proposed a novel method with high reliability and accuracy for AF detection via deep learning. The short-term Fourier transform (STFT) and stationary wavelet transform (SWT) were used to analyze ECG segments to obtain two-dimensional (2-D) matrix input suitable for deep convolutional neural networks. Then, two different deep convolutional neural network models corresponding to STFT output and SWT output were developed. Our new method did not require detection of P or R peaks, nor feature designs for classification, in contrast to existing algorithms. Finally, the performances of the two models were evaluated and compared with those of existing algorithms. Our proposed method demonstrated favorable performances on ECG segments as short as 5 s. The deep convolutional neural network using input generated by STFT, presented a sensitivity of 98.34%, specificity of 98.24% and accuracy of 98.29%. For the deep convolutional neural network using input generated by SWT, a sensitivity of 98.79%, specificity of 97.87% and accuracy of 98.63% was achieved. The proposed method using deep convolutional neural networks shows high sensitivity, specificity and accuracy, and, therefore, is a valuable tool for AF detection. Copyright © 2017 Elsevier Ltd. All rights reserved.

Deep Space Network equipment performance, reliability, and operations management information system

NASA Technical Reports Server (NTRS)

Cooper, T.; Lin, J.; Chatillon, M.

2002-01-01

The Deep Space Mission System (DSMS) Operations Program Office and the DeepSpace Network (DSN) facilities utilize the Discrepancy Reporting Management System (DRMS) to collect, process, communicate and manage data discrepancies, equipment resets, physical equipment status, and to maintain an internal Station Log. A collaborative effort development between JPL and the Canberra Deep Space Communication Complex delivered a system to support DSN Operations.

The Image of the Negro in Deep South Public School State History Texts.

ERIC Educational Resources Information Center

McLaurin, Melton

This report reviews the image portrayed of the Negro, in textbooks used in the deep South. Slavery is painted as a cordial, humane system under kindly masters and the Negro as docile and childlike. Although the treatment of the modern era is relatively more objective, the texts, on the whole, evade treatment of the Civil Rights struggle, violence,…

Wolf: What's On the Lunar Farside?

NASA Technical Reports Server (NTRS)

2008-01-01

WOLF (What's On the Lunar Farside?) is a lunar sample return mission to the South Pole-Aitken (SPA) Basin, located on the farside of the moon, seeking to answer some of the remaining questions about our solar system. Through the return and analysis of SPA samples, scientists can constrain the period of inner solar system late heavy bombardment and gain momentous knowledge of the SPA basin. WOLF provides the opportunity for mankind's progression in further understanding our solar system, its history, and unknowns surrounding the lunar farside. The orbiter will provide intermittent, direct communication between the lander and ground operations via the Deep Space Network (DSN). Received images and spectrometry will aid in real-time sample selection.

The 2008 earthquakes in the Bavarian Molasse Basin - possible relation to deep geothermics?

NASA Astrophysics Data System (ADS)

Kraft, T.; Wassermann, J.; Deichmann, N.; Stange, S.

2009-04-01

We discuss several microearthquakes of magnitude up to Ml=2.3 that occurred in the Bavarian Molasse Basin (ByM), south of Munich, Germany, in February and July 2008. The strongest event was felt by local residents. The Bavarian Earthquake catalog, which dates back to the year 1000, does list a small number of isolated earthquakes in the western part of the ByM as well as a cluster of mining induced earthquakes (Peißenberg 1962-1970, I0(MSK)=5.5). The eastern part of the ByM, including the wider surrounding of Munich, was so far considered aseismic. Due to the spatio-temporal clustering of the microearthquakes in February and July 2008 the University of Munich (LMU) and the Swiss Seismologcical Service installed a temporal network of seismological stations in the south of Munich to investigate the newly arising seismicity. First analysis of the recorded data indicate shallow source depths (~5km) for the July events. This result is supported by the fact that one of these very small earthquakes was felt by local residents. The earthquakes hypocenters are located closely to a number of deep geothermal wells of 3-4.5km depth being either in production or running productivity tests in late 2007 and early 2008. Therefore, the 2008 seimicity might represent a case of induced seimicity related to the injection or withdrawal of water from the hydrothermal aquifer. Due to the lack of high quality recordings of a denser seismic monitoring network in the source area it is not possible to resolve details of the processes behind the 2008 seismicity. Therefore, a definite answer to the question if the earthquakes are related the deep geothermal projects or not can not be given at present. However, a number of recent well-studied cases have proved that earthquakes can also happen in depths much shallower than 5km, and that small changes of the hydrological conditions at depth are sufficient to trigger seismicity. Therefore, a detailed understanding of the causative processes behind the 2008 seismicity in the ByM is of paramount importance to hazard assessment and mitigation associated with similar geothermal projects underway elsewhere. A close cooperation of operators and developers of geothermal projects with earthquake science has proved to be very beneficial in the development of the Hot-Dry-Rock technique and is also highly desirable in developing strategies for the save geothermal use of deep hydrothermal aquifers.

Prediction of properties of wheat dough using intelligent deep belief networks

NASA Astrophysics Data System (ADS)

Guha, Paramita; Bhatnagar, Taru; Pal, Ishan; Kamboj, Uma; Mishra, Sunita

2017-11-01

In this paper, the rheological and chemical properties of wheat dough are predicted using deep belief networks. Wheat grains are stored at controlled environmental conditions. The internal parameters of grains viz., protein, fat, carbohydrates, moisture, ash are determined using standard chemical analysis and viscosity of the dough is measured using Rheometer. Here, fat, carbohydrates, moisture, ash and temperature are considered as inputs whereas protein and viscosity are chosen as outputs. The prediction algorithm is developed using deep neural network where each layer is trained greedily using restricted Boltzmann machine (RBM) networks. The overall network is finally fine-tuned using standard neural network technique. In most literature, it has been found that fine-tuning is done using back-propagation technique. In this paper, a new algorithm is proposed in which each layer is tuned using RBM and the final network is fine-tuned using deep neural network (DNN). It has been observed that with the proposed algorithm, errors between the actual and predicted outputs are less compared to the conventional algorithm. Hence, the given network can be considered as beneficial as it predicts the outputs more accurately. Numerical results along with discussions are presented.

Training Deep Spiking Neural Networks Using Backpropagation.

PubMed

Lee, Jun Haeng; Delbruck, Tobi; Pfeiffer, Michael

2016-01-01

Deep spiking neural networks (SNNs) hold the potential for improving the latency and energy efficiency of deep neural networks through data-driven event-based computation. However, training such networks is difficult due to the non-differentiable nature of spike events. In this paper, we introduce a novel technique, which treats the membrane potentials of spiking neurons as differentiable signals, where discontinuities at spike times are considered as noise. This enables an error backpropagation mechanism for deep SNNs that follows the same principles as in conventional deep networks, but works directly on spike signals and membrane potentials. Compared with previous methods relying on indirect training and conversion, our technique has the potential to capture the statistics of spikes more precisely. We evaluate the proposed framework on artificially generated events from the original MNIST handwritten digit benchmark, and also on the N-MNIST benchmark recorded with an event-based dynamic vision sensor, in which the proposed method reduces the error rate by a factor of more than three compared to the best previous SNN, and also achieves a higher accuracy than a conventional convolutional neural network (CNN) trained and tested on the same data. We demonstrate in the context of the MNIST task that thanks to their event-driven operation, deep SNNs (both fully connected and convolutional) trained with our method achieve accuracy equivalent with conventional neural networks. In the N-MNIST example, equivalent accuracy is achieved with about five times fewer computational operations.

The deep space network, volume 15

NASA Technical Reports Server (NTRS)

1973-01-01

The DSN progress is reported in flight project support, TDA research and technology, network engineering, hardware and software implementation, and operations. Topics discussed include: DSN functions and facilities, planetary flight projects, tracking and ground-based navigation, communications, data processing, network control system, and deep space stations.

The Deep Space Network, volume 39

NASA Technical Reports Server (NTRS)

1977-01-01

The functions, facilities, and capabilities of the Deep Space Network and its support of the Pioneer, Helios, and Viking missions are described. Progress in tracking and data acquisition research and technology, network engineering and modifications, as well as hardware and software implementation and operations are reported.

Deep space network Mark 4A description

NASA Technical Reports Server (NTRS)

Wallace, R. J.; Burt, R. W.

1986-01-01

The general system configuration for the Mark 4A Deep Space Network is described. The arrangement and complement of antennas at the communications complexes and subsystem equipment at the signal processing centers are described. A description of the Network Operations Control Center is also presented.

Toolkits and Libraries for Deep Learning.

PubMed

Erickson, Bradley J; Korfiatis, Panagiotis; Akkus, Zeynettin; Kline, Timothy; Philbrick, Kenneth

2017-08-01

Deep learning is an important new area of machine learning which encompasses a wide range of neural network architectures designed to complete various tasks. In the medical imaging domain, example tasks include organ segmentation, lesion detection, and tumor classification. The most popular network architecture for deep learning for images is the convolutional neural network (CNN). Whereas traditional machine learning requires determination and calculation of features from which the algorithm learns, deep learning approaches learn the important features as well as the proper weighting of those features to make predictions for new data. In this paper, we will describe some of the libraries and tools that are available to aid in the construction and efficient execution of deep learning as applied to medical images.

Future Mission Trends and their Implications for the Deep Space Network

NASA Technical Reports Server (NTRS)

Abraham, Douglas S.

2006-01-01

This viewgraph presentation discusses the direction of future missions and it's significance to the Deep Space Network. The topics include: 1) The Deep Space Network (DSN); 2) Past Missions Driving DSN Evolution; 3) The Changing Mission Paradigm; 4) Assessing Future Mission Needs; 5) Link Support Trends; 6) Downlink Rate Trends; 7) Uplink Rate Trends; 8) End-to-End Link Difficulty Trends; 9) Summary: Future Mission Trend Drivers; and 10) Conclusion: Implications for the DSN.

The deep space network, volume 6

NASA Technical Reports Server (NTRS)

1971-01-01

Progress on Deep Space Network (DSN) supporting research and technology is presented, together with advanced development and engineering, implementation, and DSN operations of flight projects. The DSN is described. Interplanetary and planetary flight projects and radio science experiments are discussed. Tracking and navigational accuracy analysis, communications systems and elements research, and supporting research are considered. Development of the ground communications and deep space instrumentation facilities is also presented. Network allocation schedules and angle tracking and test development are included.

The Deep Space Network

NASA Technical Reports Server (NTRS)

1977-01-01

The various systems and subsystems are discussed for the Deep Space Network (DSN). A description of the DSN is presented along with mission support, program planning, facility engineering, implementation and operations.

Hourly air pollution concentrations and their important predictors over Houston, Texas using deep neural networks: case study of DISCOVER-AQ time period

NASA Astrophysics Data System (ADS)

Eslami, E.; Choi, Y.; Roy, A.

2017-12-01

Air quality forecasting carried out by chemical transport models often show significant error. This study uses a deep-learning approach over the Houston-Galveston-Brazoria (HGB) area to overcome this forecasting challenge, for the DISCOVER-AQ period (September 2013). Two approaches, deep neural network (DNN) using a Multi-Layer Perceptron (MLP) and Restricted Boltzmann Machine (RBM) were utilized. The proposed approaches analyzed input data by identifying features abstracted from its previous layer using a stepwise method. The approaches predicted hourly ozone and PM in September 2013 using several predictors of prior three days, including wind fields, temperature, relative humidity, cloud fraction, precipitation along with PM, ozone, and NOx concentrations. Model-measurement comparisons for available monitoring sites reported Indexes of Agreement (IOA) of around 0.95 for both DNN and RBM. A standard artificial neural network (ANN) (IOA=0.90) with similar architecture showed poorer performance than the deep networks, clearly demonstrating the superiority of the deep approaches. Additionally, each network (both deep and standard) performed significantly better than a previous CMAQ study, which showed an IOA of less than 0.80. The most influential input variables were identified using their associated weights, which represented the sensitivity of ozone to input parameters. The results indicate deep learning approaches can achieve more accurate ozone forecasting and identify the important input variables for ozone predictions in metropolitan areas.

De novo peptide sequencing by deep learning

PubMed Central

Tran, Ngoc Hieu; Zhang, Xianglilan; Xin, Lei; Shan, Baozhen; Li, Ming

2017-01-01

De novo peptide sequencing from tandem MS data is the key technology in proteomics for the characterization of proteins, especially for new sequences, such as mAbs. In this study, we propose a deep neural network model, DeepNovo, for de novo peptide sequencing. DeepNovo architecture combines recent advances in convolutional neural networks and recurrent neural networks to learn features of tandem mass spectra, fragment ions, and sequence patterns of peptides. The networks are further integrated with local dynamic programming to solve the complex optimization task of de novo sequencing. We evaluated the method on a wide variety of species and found that DeepNovo considerably outperformed state of the art methods, achieving 7.7–22.9% higher accuracy at the amino acid level and 38.1–64.0% higher accuracy at the peptide level. We further used DeepNovo to automatically reconstruct the complete sequences of antibody light and heavy chains of mouse, achieving 97.5–100% coverage and 97.2–99.5% accuracy, without assisting databases. Moreover, DeepNovo is retrainable to adapt to any sources of data and provides a complete end-to-end training and prediction solution to the de novo sequencing problem. Not only does our study extend the deep learning revolution to a new field, but it also shows an innovative approach in solving optimization problems by using deep learning and dynamic programming. PMID:28720701

Major technological innovations introduced in the large antennas of the Deep Space Network

NASA Technical Reports Server (NTRS)

Imbriale, W. A.

2002-01-01

The NASA Deep Space Network (DSN) is the largest and most sensitive scientific, telecommunications and radio navigation network in the world. Its principal responsibilities are to provide communications, tracking, and science services to most of the world's spacecraft that travel beyond low Earth orbit. The network consists of three Deep Space Communications Complexes. Each of the three complexes consists of multiple large antennas equipped with ultra sensitive receiving systems. A centralized Signal Processing Center (SPC) remotely controls the antennas, generates and transmits spacecraft commands, and receives and processes the spacecraft telemetry.

7.3 Communications and Navigation

NASA Technical Reports Server (NTRS)

Manning, Rob

2005-01-01

This presentation gives an overview of the networks NASA currently uses to support space communications and navigation, and the requirements for supporting future deep space missions, including manned lunar and Mars missions. The presentation addresses the Space Network, Deep Space Network, and Ground Network, why new support systems are needed, and the potential for catastrophic failure of aging antennas. Space communications and navigation are considered during Aerocapture, Entry, Descent and Landing (AEDL) only in order to precisely position, track and interact with the spacecraft at its destination (moon, Mars and Earth return) arrival. The presentation recommends a combined optical/radio frequency strategy for deep space communications.

Hydrogen peroxide in deep waters from the Mediterranean Sea, South Atlantic and South Pacific Oceans

NASA Astrophysics Data System (ADS)

Hopwood, Mark J.; Rapp, Insa; Schlosser, Christian; Achterberg, Eric P.

2017-03-01

Hydrogen peroxide (H2O2) is present ubiquitously in marine surface waters where it is a reactive intermediate in the cycling of many trace elements. Photochemical processes are considered the dominant natural H2O2 source, yet cannot explain nanomolar H2O2 concentrations below the photic zone. Here, we determined the concentration of H2O2 in full depth profiles across three ocean basins (Mediterranean Sea, South Atlantic and South Pacific Oceans). To determine the accuracy of H2O2 measurements in the deep ocean we also re-assessed the contribution of interfering species to ‘apparent H2O2’, as analysed by the luminol based chemiluminescence technique. Within the vicinity of coastal oxygen minimum zones, accurate measurement of H2O2 was not possible due to interference from Fe(II). Offshore, in deep (>1000 m) waters H2O2 concentrations ranged from 0.25 ± 0.27 nM (Mediterranean, Balearics-Algeria) to 2.9 ± 2.2 nM (Mediterranean, Corsica-France). Our results indicate that a dark, pelagic H2O2 production mechanism must occur throughout the deep ocean. A bacterial source of H2O2 is the most likely origin and we show that this source is likely sufficient to account for all of the observed H2O2 in the deep ocean.

Hydrogen peroxide in deep waters from the Mediterranean Sea, South Atlantic and South Pacific Oceans

PubMed Central

Hopwood, Mark J.; Rapp, Insa; Schlosser, Christian; Achterberg, Eric P.

2017-01-01

Hydrogen peroxide (H2O2) is present ubiquitously in marine surface waters where it is a reactive intermediate in the cycling of many trace elements. Photochemical processes are considered the dominant natural H2O2 source, yet cannot explain nanomolar H2O2 concentrations below the photic zone. Here, we determined the concentration of H2O2 in full depth profiles across three ocean basins (Mediterranean Sea, South Atlantic and South Pacific Oceans). To determine the accuracy of H2O2 measurements in the deep ocean we also re-assessed the contribution of interfering species to ‘apparent H2O2’, as analysed by the luminol based chemiluminescence technique. Within the vicinity of coastal oxygen minimum zones, accurate measurement of H2O2 was not possible due to interference from Fe(II). Offshore, in deep (>1000 m) waters H2O2 concentrations ranged from 0.25 ± 0.27 nM (Mediterranean, Balearics-Algeria) to 2.9 ± 2.2 nM (Mediterranean, Corsica-France). Our results indicate that a dark, pelagic H2O2 production mechanism must occur throughout the deep ocean. A bacterial source of H2O2 is the most likely origin and we show that this source is likely sufficient to account for all of the observed H2O2 in the deep ocean. PMID:28266529

Evolution of South Atlantic density and chemical stratification across the last deglaciation

PubMed Central

Skinner, Luke C.; Peck, Victoria L.; Kender, Sev; Elderfield, Henry; Waelbroeck, Claire; Hodell, David A.

2016-01-01

Explanations of the glacial–interglacial variations in atmospheric pCO2 invoke a significant role for the deep ocean in the storage of CO2. Deep-ocean density stratification has been proposed as a mechanism to promote the storage of CO2 in the deep ocean during glacial times. A wealth of proxy data supports the presence of a “chemical divide” between intermediate and deep water in the glacial Atlantic Ocean, which indirectly points to an increase in deep-ocean density stratification. However, direct observational evidence of changes in the primary controls of ocean density stratification, i.e., temperature and salinity, remain scarce. Here, we use Mg/Ca-derived seawater temperature and salinity estimates determined from temperature-corrected δ18O measurements on the benthic foraminifer Uvigerina spp. from deep and intermediate water-depth marine sediment cores to reconstruct the changes in density of sub-Antarctic South Atlantic water masses over the last deglaciation (i.e., 22–2 ka before present). We find that a major breakdown in the physical density stratification significantly lags the breakdown of the deep-intermediate chemical divide, as indicated by the chemical tracers of benthic foraminifer δ13C and foraminifer/coral 14C. Our results indicate that chemical destratification likely resulted in the first rise in atmospheric pCO2, whereas the density destratification of the deep South Atlantic lags the second rise in atmospheric pCO2 during the late deglacial period. Our findings emphasize that the physical and chemical destratification of the ocean are not as tightly coupled as generally assumed. PMID:26729858

Evolution of South Atlantic density and chemical stratification across the last deglaciation.

PubMed

Roberts, Jenny; Gottschalk, Julia; Skinner, Luke C; Peck, Victoria L; Kender, Sev; Elderfield, Henry; Waelbroeck, Claire; Vázquez Riveiros, Natalia; Hodell, David A

2016-01-19

Explanations of the glacial-interglacial variations in atmospheric pCO2 invoke a significant role for the deep ocean in the storage of CO2. Deep-ocean density stratification has been proposed as a mechanism to promote the storage of CO2 in the deep ocean during glacial times. A wealth of proxy data supports the presence of a "chemical divide" between intermediate and deep water in the glacial Atlantic Ocean, which indirectly points to an increase in deep-ocean density stratification. However, direct observational evidence of changes in the primary controls of ocean density stratification, i.e., temperature and salinity, remain scarce. Here, we use Mg/Ca-derived seawater temperature and salinity estimates determined from temperature-corrected δ(18)O measurements on the benthic foraminifer Uvigerina spp. from deep and intermediate water-depth marine sediment cores to reconstruct the changes in density of sub-Antarctic South Atlantic water masses over the last deglaciation (i.e., 22-2 ka before present). We find that a major breakdown in the physical density stratification significantly lags the breakdown of the deep-intermediate chemical divide, as indicated by the chemical tracers of benthic foraminifer δ(13)C and foraminifer/coral (14)C. Our results indicate that chemical destratification likely resulted in the first rise in atmospheric pCO2, whereas the density destratification of the deep South Atlantic lags the second rise in atmospheric pCO2 during the late deglacial period. Our findings emphasize that the physical and chemical destratification of the ocean are not as tightly coupled as generally assumed.

Considerations on communications network protocols in deep space

NASA Technical Reports Server (NTRS)

Clare, L. P.; Agre, J. R.; Yan, T.

2001-01-01

Communications supporting deep space missions impose numerous unique constraints that impact the architectural choices made for cost-effectiveness. We are entering the era where networks that exist in deep space are needed to support planetary exploration. Cost-effective performance will require a balanced integration of applicable widely used standard protocols with new and innovative designs.

Saver.net lidar network in southern South America

NASA Astrophysics Data System (ADS)

Ristori, Pablo; Otero, Lidia; Jin, Yoshitaka; Barja, Boris; Shimizu, Atsushi; Barbero, Albane; Salvador, Jacobo; Bali, Juan Lucas; Herrera, Milagros; Etala, Paula; Acquesta, Alejandro; Quel, Eduardo; Sugimoto, Nobuo; Mizuno, Akira

2018-04-01

The South American Environmental Risk Management Network (SAVER-Net) is an instrumentation network, mainly composed by lidars, to provide real-time information for atmospheric hazards and risk management purposes in South America. This lidar network have been developed since 2012 and all its sampling points are expected to be fully implemented by 2017. This paper describes the network's status and configuration, the data acquisition and processing scheme (protocols and data levels), as well as some aspects of the scientific networking in Latin American Lidar Network (LALINET). Similarly, the paper lays out future plans on the operation and integration to major international collaborative efforts.

Quantitative phase microscopy using deep neural networks

NASA Astrophysics Data System (ADS)

Li, Shuai; Sinha, Ayan; Lee, Justin; Barbastathis, George

2018-02-01

Deep learning has been proven to achieve ground-breaking accuracy in various tasks. In this paper, we implemented a deep neural network (DNN) to achieve phase retrieval in a wide-field microscope. Our DNN utilized the residual neural network (ResNet) architecture and was trained using the data generated by a phase SLM. The results showed that our DNN was able to reconstruct the profile of the phase target qualitatively. In the meantime, large error still existed, which indicated that our approach still need to be improved.

Breakup of last glacial deep stratification in the South Pacific.

PubMed

Basak, Chandranath; Fröllje, Henning; Lamy, Frank; Gersonde, Rainer; Benz, Verena; Anderson, Robert F; Molina-Kescher, Mario; Pahnke, Katharina

2018-02-23

Stratification of the deep Southern Ocean during the Last Glacial Maximum is thought to have facilitated carbon storage and subsequent release during the deglaciation as stratification broke down, contributing to atmospheric CO 2 rise. Here, we present neodymium isotope evidence from deep to abyssal waters in the South Pacific that confirms stratification of the deepwater column during the Last Glacial Maximum. The results indicate a glacial northward expansion of Ross Sea Bottom Water and a Southern Hemisphere climate trigger for the deglacial breakup of deep stratification. It highlights the important role of abyssal waters in sustaining a deep glacial carbon reservoir and Southern Hemisphere climate change as a prerequisite for the destabilization of the water column and hence the deglacial release of sequestered CO 2 through upwelling. Copyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works.

Classifying the molecular functions of Rab GTPases in membrane trafficking using deep convolutional neural networks.

PubMed

Le, Nguyen-Quoc-Khanh; Ho, Quang-Thai; Ou, Yu-Yen

2018-06-13

Deep learning has been increasingly used to solve a number of problems with state-of-the-art performance in a wide variety of fields. In biology, deep learning can be applied to reduce feature extraction time and achieve high levels of performance. In our present work, we apply deep learning via two-dimensional convolutional neural networks and position-specific scoring matrices to classify Rab protein molecules, which are main regulators in membrane trafficking for transferring proteins and other macromolecules throughout the cell. The functional loss of specific Rab molecular functions has been implicated in a variety of human diseases, e.g., choroideremia, intellectual disabilities, cancer. Therefore, creating a precise model for classifying Rabs is crucial in helping biologists understand the molecular functions of Rabs and design drug targets according to such specific human disease information. We constructed a robust deep neural network for classifying Rabs that achieved an accuracy of 99%, 99.5%, 96.3%, and 97.6% for each of four specific molecular functions. Our approach demonstrates superior performance to traditional artificial neural networks. Therefore, from our proposed study, we provide both an effective tool for classifying Rab proteins and a basis for further research that can improve the performance of biological modeling using deep neural networks. Copyright © 2018 Elsevier Inc. All rights reserved.

Towards deep learning with segregated dendrites

PubMed Central

Guerguiev, Jordan; Lillicrap, Timothy P

2017-01-01

Deep learning has led to significant advances in artificial intelligence, in part, by adopting strategies motivated by neurophysiology. However, it is unclear whether deep learning could occur in the real brain. Here, we show that a deep learning algorithm that utilizes multi-compartment neurons might help us to understand how the neocortex optimizes cost functions. Like neocortical pyramidal neurons, neurons in our model receive sensory information and higher-order feedback in electrotonically segregated compartments. Thanks to this segregation, neurons in different layers of the network can coordinate synaptic weight updates. As a result, the network learns to categorize images better than a single layer network. Furthermore, we show that our algorithm takes advantage of multilayer architectures to identify useful higher-order representations—the hallmark of deep learning. This work demonstrates that deep learning can be achieved using segregated dendritic compartments, which may help to explain the morphology of neocortical pyramidal neurons. PMID:29205151

Towards deep learning with segregated dendrites.

PubMed

Guerguiev, Jordan; Lillicrap, Timothy P; Richards, Blake A

2017-12-05

Deep learning has led to significant advances in artificial intelligence, in part, by adopting strategies motivated by neurophysiology. However, it is unclear whether deep learning could occur in the real brain. Here, we show that a deep learning algorithm that utilizes multi-compartment neurons might help us to understand how the neocortex optimizes cost functions. Like neocortical pyramidal neurons, neurons in our model receive sensory information and higher-order feedback in electrotonically segregated compartments. Thanks to this segregation, neurons in different layers of the network can coordinate synaptic weight updates. As a result, the network learns to categorize images better than a single layer network. Furthermore, we show that our algorithm takes advantage of multilayer architectures to identify useful higher-order representations-the hallmark of deep learning. This work demonstrates that deep learning can be achieved using segregated dendritic compartments, which may help to explain the morphology of neocortical pyramidal neurons.

DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network.

PubMed

Katzman, Jared L; Shaham, Uri; Cloninger, Alexander; Bates, Jonathan; Jiang, Tingting; Kluger, Yuval

2018-02-26

Medical practitioners use survival models to explore and understand the relationships between patients' covariates (e.g. clinical and genetic features) and the effectiveness of various treatment options. Standard survival models like the linear Cox proportional hazards model require extensive feature engineering or prior medical knowledge to model treatment interaction at an individual level. While nonlinear survival methods, such as neural networks and survival forests, can inherently model these high-level interaction terms, they have yet to be shown as effective treatment recommender systems. We introduce DeepSurv, a Cox proportional hazards deep neural network and state-of-the-art survival method for modeling interactions between a patient's covariates and treatment effectiveness in order to provide personalized treatment recommendations. We perform a number of experiments training DeepSurv on simulated and real survival data. We demonstrate that DeepSurv performs as well as or better than other state-of-the-art survival models and validate that DeepSurv successfully models increasingly complex relationships between a patient's covariates and their risk of failure. We then show how DeepSurv models the relationship between a patient's features and effectiveness of different treatment options to show how DeepSurv can be used to provide individual treatment recommendations. Finally, we train DeepSurv on real clinical studies to demonstrate how it's personalized treatment recommendations would increase the survival time of a set of patients. The predictive and modeling capabilities of DeepSurv will enable medical researchers to use deep neural networks as a tool in their exploration, understanding, and prediction of the effects of a patient's characteristics on their risk of failure.

The Network Information Management System (NIMS) in the Deep Space Network

NASA Technical Reports Server (NTRS)

Wales, K. J.

1983-01-01

In an effort to better manage enormous amounts of administrative, engineering, and management data that is distributed worldwide, a study was conducted which identified the need for a network support system. The Network Information Management System (NIMS) will provide the Deep Space Network with the tools to provide an easily accessible source of valid information to support management activities and provide a more cost-effective method of acquiring, maintaining, and retrieval data.

Robustness of spiking Deep Belief Networks to noise and reduced bit precision of neuro-inspired hardware platforms.

PubMed

Stromatias, Evangelos; Neil, Daniel; Pfeiffer, Michael; Galluppi, Francesco; Furber, Steve B; Liu, Shih-Chii

2015-01-01

Increasingly large deep learning architectures, such as Deep Belief Networks (DBNs) are the focus of current machine learning research and achieve state-of-the-art results in different domains. However, both training and execution of large-scale Deep Networks require vast computing resources, leading to high power requirements and communication overheads. The on-going work on design and construction of spike-based hardware platforms offers an alternative for running deep neural networks with significantly lower power consumption, but has to overcome hardware limitations in terms of noise and limited weight precision, as well as noise inherent in the sensor signal. This article investigates how such hardware constraints impact the performance of spiking neural network implementations of DBNs. In particular, the influence of limited bit precision during execution and training, and the impact of silicon mismatch in the synaptic weight parameters of custom hybrid VLSI implementations is studied. Furthermore, the network performance of spiking DBNs is characterized with regard to noise in the spiking input signal. Our results demonstrate that spiking DBNs can tolerate very low levels of hardware bit precision down to almost two bits, and show that their performance can be improved by at least 30% through an adapted training mechanism that takes the bit precision of the target platform into account. Spiking DBNs thus present an important use-case for large-scale hybrid analog-digital or digital neuromorphic platforms such as SpiNNaker, which can execute large but precision-constrained deep networks in real time.

Robustness of spiking Deep Belief Networks to noise and reduced bit precision of neuro-inspired hardware platforms

PubMed Central

Stromatias, Evangelos; Neil, Daniel; Pfeiffer, Michael; Galluppi, Francesco; Furber, Steve B.; Liu, Shih-Chii

2015-01-01

Increasingly large deep learning architectures, such as Deep Belief Networks (DBNs) are the focus of current machine learning research and achieve state-of-the-art results in different domains. However, both training and execution of large-scale Deep Networks require vast computing resources, leading to high power requirements and communication overheads. The on-going work on design and construction of spike-based hardware platforms offers an alternative for running deep neural networks with significantly lower power consumption, but has to overcome hardware limitations in terms of noise and limited weight precision, as well as noise inherent in the sensor signal. This article investigates how such hardware constraints impact the performance of spiking neural network implementations of DBNs. In particular, the influence of limited bit precision during execution and training, and the impact of silicon mismatch in the synaptic weight parameters of custom hybrid VLSI implementations is studied. Furthermore, the network performance of spiking DBNs is characterized with regard to noise in the spiking input signal. Our results demonstrate that spiking DBNs can tolerate very low levels of hardware bit precision down to almost two bits, and show that their performance can be improved by at least 30% through an adapted training mechanism that takes the bit precision of the target platform into account. Spiking DBNs thus present an important use-case for large-scale hybrid analog-digital or digital neuromorphic platforms such as SpiNNaker, which can execute large but precision-constrained deep networks in real time. PMID:26217169

Radiocarbon constraints on the extent and evolution of the South Pacific glacial carbon pool

PubMed Central

Ronge, T. A.; Tiedemann, R.; Lamy, F.; Köhler, P.; Alloway, B. V.; De Pol-Holz, R.; Pahnke, K.; Southon, J.; Wacker, L.

2016-01-01

During the last deglaciation, the opposing patterns of atmospheric CO2 and radiocarbon activities (Δ14C) suggest the release of 14C-depleted CO2 from old carbon reservoirs. Although evidences point to the deep Pacific as a major reservoir of this 14C-depleted carbon, its extent and evolution still need to be constrained. Here we use sediment cores retrieved along a South Pacific transect to reconstruct the spatio-temporal evolution of Δ14C over the last 30,000 years. In ∼2,500–3,600 m water depth, we find 14C-depleted deep waters with a maximum glacial offset to atmospheric 14C (ΔΔ14C=−1,000‰). Using a box model, we test the hypothesis that these low values might have been caused by an interaction of aging and hydrothermal CO2 influx. We observe a rejuvenation of circumpolar deep waters synchronous and potentially contributing to the initial deglacial rise in atmospheric CO2. These findings constrain parts of the glacial carbon pool to the deep South Pacific. PMID:27157845

Antifouling potentials of eight deep-sea-derived fungi from the South China Sea.

PubMed

Zhang, Xiao-Yong; Xu, Xin-Ya; Peng, Jiang; Ma, Chun-Feng; Nong, Xu-Hua; Bao, Jie; Zhang, Guang-Zhao; Qi, Shu-Hua

2014-04-01

Marine-derived microbial secondary metabolites are promising potential sources of nontoxic antifouling agents. The search for environmentally friendly and low-toxic antifouling components guided us to investigate the antifouling potentials of eight novel fungal isolates from deep-sea sediments of the South China Sea. Sixteen crude ethyl acetate extracts of the eight fungal isolates showed distinct antibacterial activity against three marine bacteria (Loktanella hongkongensis UST950701-009, Micrococcus luteus UST950701-006 and Pseudoalteromonas piscida UST010620-005), or significant antilarval activity against larval settlement of bryozoan Bugula neritina. Furthermore, the extract of Aspergillus westerdijkiae DFFSCS013 displayed strong antifouling activity in a field trial lasting 4 months. By further bioassay-guided isolation, five antifouling alkaloids including brevianamide F, circumdatin F and L, notoamide C, and 5-chlorosclerotiamide were isolated from the extract of A. westerdijkiae DFFSCS013. This is the first report about the antifouling potentials of metabolites of the deep-sea-derived fungi from the South China Sea, and the first stage towards the development of non- or low-toxic antifouling agents from deep-sea-derived fungi.

Deep learning in bioinformatics.

PubMed

Min, Seonwoo; Lee, Byunghan; Yoon, Sungroh

2017-09-01

In the era of big data, transformation of biomedical big data into valuable knowledge has been one of the most important challenges in bioinformatics. Deep learning has advanced rapidly since the early 2000s and now demonstrates state-of-the-art performance in various fields. Accordingly, application of deep learning in bioinformatics to gain insight from data has been emphasized in both academia and industry. Here, we review deep learning in bioinformatics, presenting examples of current research. To provide a useful and comprehensive perspective, we categorize research both by the bioinformatics domain (i.e. omics, biomedical imaging, biomedical signal processing) and deep learning architecture (i.e. deep neural networks, convolutional neural networks, recurrent neural networks, emergent architectures) and present brief descriptions of each study. Additionally, we discuss theoretical and practical issues of deep learning in bioinformatics and suggest future research directions. We believe that this review will provide valuable insights and serve as a starting point for researchers to apply deep learning approaches in their bioinformatics studies. © The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

The Deep Space Network: The challenges of the next 20 years - The 21st century

NASA Technical Reports Server (NTRS)

Dumas, L. N.; Edwards, C. D.; Hall, J. R.; Posner, E. C.

1990-01-01

The Deep Space Network (DSN) has been the radio navigation and communications link between NASA's lunar and deep space missions for 30 years. In this paper, new mission opportunities over the next 20 years are discussed. The system design drivers and the DSN architectural concepts for those challenges are briefly considered.

Drilling a deep geologic test well at Hilton Head Island, South Carolina

USGS Publications Warehouse

Schultz, Arthur P.; Seefelt, Ellen L.

2011-01-01

The U.S. Geological Survey, in cooperation with the South Carolina Department of Health and Environmental Control (SCDHEC), is drilling a deep geologic test well at Hilton Head Island, S.C. The test well is scheduled to run between mid-March and early May 2011. When completed, the well will be about 1,000 feet deep. The purpose of this test well is to gain knowledge about the regional-scale Floridan aquifer, an important source of groundwater in the Hilton Head area. Also, cores obtained during drilling will enable geologists to study the last 60 million years of Earth history in this area.

Deep Convolutional Framelet Denosing for Low-Dose CT via Wavelet Residual Network.

PubMed

Kang, Eunhee; Chang, Won; Yoo, Jaejun; Ye, Jong Chul

2018-06-01

Model-based iterative reconstruction algorithms for low-dose X-ray computed tomography (CT) are computationally expensive. To address this problem, we recently proposed a deep convolutional neural network (CNN) for low-dose X-ray CT and won the second place in 2016 AAPM Low-Dose CT Grand Challenge. However, some of the textures were not fully recovered. To address this problem, here we propose a novel framelet-based denoising algorithm using wavelet residual network which synergistically combines the expressive power of deep learning and the performance guarantee from the framelet-based denoising algorithms. The new algorithms were inspired by the recent interpretation of the deep CNN as a cascaded convolution framelet signal representation. Extensive experimental results confirm that the proposed networks have significantly improved performance and preserve the detail texture of the original images.

[Diversity of culturable sulfur-oxidizing bacteria in deep-sea hydrothermal vent environments of the South Atlantic].

PubMed

Xu, Hongxiu; Jiang, Lijing; Li, Shaoneng; Zhong, Tianhua; Lai, Qiliang; Shao, Zongze

2016-01-04

To investigate the diversity of culturable sulfur-oxidizing bacteria in hydrothermal vent environments of the South Atlantic, and analyze their characteristics of sulfur oxidation. We enriched and isolated sulfur-oxidizing bacteria from hydrothermal vent samples collected from the South Atlantic. The microbial diversity in enrichment cultures was analyzed using the Denatural Gradient Gel Electrophoresis method. Sulfur-oxidizing characteristics of the isolates was further studied by using ion chromatography. A total of 48 isolates were obtained from the deep-sea hydrothermal vent samples, which belonged to 23 genera and mainly grouped into alpha-Proteobacteria (58.3%), Actinobacteria (22.9%) and gama-Proteobacteria (18.8%). Among them, the genus Thalassospira, Martelella and Microbacterium were dominant. About 60% of the isolates exibited sulfur-oxidizing ability and strain L6M1-5 had a higher sulfur oxidation rate by comparison analysis. The diversity of sulfur-oxidizing bacteria in hydrothermal environments of the South Atlantic was reported for the first time based on culture-dependent methods. The result will help understand the biogechemical process of sulfur compounds in the deep-sea hydrothermal environments.

Using Deep Slow Slip in New Zealand to Constrain Slip Partitioning

NASA Astrophysics Data System (ADS)

Bartlow, N. M.; Wallace, L. M.

2016-12-01

Underneath New Zealand's North Island, the Pacific plate subducts obliquely beneath the Australian plate. Just to the south, subduction ceases and the plate boundary transitions to the mainly strike-slip, steeply dipping Alpine fault that runs along the South Island. In the region of the southern North Island, the relative plate motion has significant components of both convergence and along strike motion, and slip is partitioned between the main Hikurangi subduction interface and a series of shallower strike-slip faults running thurough the North Island (Wallace and Beavan, GRL, 2010). This region also hosts deep ( 50 km), long duration ( 1 year) slow slip events (SSEs). From early 2013 to early 2016, continuous GPS stations maintained by GeoNet in this region recorded two such deep SSEs on the Hikurangi megathrust. The first SSE occurred on the Kapiti patch, just southwest of the North Island coast. SSEs previous occurred here in 2003 and 2008 (Wallace and Beavan, JGR, 2010). The 2014 Kapiti SSE is unique because it was rapidly decelerated following increased normal stress (clamping) caused by a nearby M 6.3 earthquake (Wallace et al., GRL, 2014). However, GPS data indicates that slip did not stop entirely, and soon after the Manawatu slow slip patch just to the northeast ruptured in another SSE. This patch previously had large SSEs in 2004/2005 and 2010/2011. Given the previous repeat interval of 5.5 years, the 2014/2015 Manawatu SSE is early; however, the record is very short. Here we show Network Inversion Filter derived models of slow slip for the various phases of the Kapiti and Manawatu SSEs, which indicate a possible continuous migration of slip from the Kapiti SSE patch to the Manawatu SSE patch, and we quantify the shear stress increase on the Manawatu patch after the Kapiti SSE. Additionally, we explore allowing the Network Inversion Filter to vary the direction of slip on the plate interface to better fit the data. We estimate how much of the strike-slip and dip-slip components of the relative plate motion are being accommodated by the main thrust interface, and infer how much slip is being accommodated by the strike-slip faults and forearc rotation. We compare our results to those from prior block models of inter-SSE data (Wallace et al., G3, 2009) and explore the implications for seismic hazard assessment in this region.

The Deep Space Network

NASA Technical Reports Server (NTRS)

1988-01-01

The Deep Space Network (DSN) is the largest and most sensitive scientific telecommunications and radio navigation network in the world. Its principal responsibilities are to support unmanned interplanetary spacecraft missions and to support radio and radar astronomy observations in the exploration of the solar system and the universe. The DSN facilities and capabilities as of January 1988 are described.

The deep space network. [tracking and communication support for space probes

NASA Technical Reports Server (NTRS)

1974-01-01

The objectives, functions, and organization of the deep space network are summarized. Progress in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations is reported. Interface support for the Mariner Venus Mercury 1973 flight and Pioneer 10 and 11 missions is included.

The deep space network, volume 12

NASA Technical Reports Server (NTRS)

1972-01-01

Progress in the development of the DSN is reported along with TDA research and technology, network engineering, hardware, and software implementation. Included are descriptions of the DSN function and facilities, Helios mission support, Mariner Venus/Mercury 1973 mission support, Viking mission support, tracking and ground-based navigation, communications, network control and data processing, and deep space stations.

Performance of Deep and Shallow Neural Networks, the Universal Approximation Theorem, Activity Cliffs, and QSAR.

PubMed

Winkler, David A; Le, Tu C

2017-01-01

Neural networks have generated valuable Quantitative Structure-Activity/Property Relationships (QSAR/QSPR) models for a wide variety of small molecules and materials properties. They have grown in sophistication and many of their initial problems have been overcome by modern mathematical techniques. QSAR studies have almost always used so-called "shallow" neural networks in which there is a single hidden layer between the input and output layers. Recently, a new and potentially paradigm-shifting type of neural network based on Deep Learning has appeared. Deep learning methods have generated impressive improvements in image and voice recognition, and are now being applied to QSAR and QSAR modelling. This paper describes the differences in approach between deep and shallow neural networks, compares their abilities to predict the properties of test sets for 15 large drug data sets (the kaggle set), discusses the results in terms of the Universal Approximation theorem for neural networks, and describes how DNN may ameliorate or remove troublesome "activity cliffs" in QSAR data sets. © 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim.

Regional ground-water discharge to large streams in the upper coastal plain of South Carolina and parts of North Carolina and Georgia

USGS Publications Warehouse

Aucott, W.R.; Meadows, R.S.; Patterson, G.G.

1987-01-01

Base flow was computed to estimate discharge from regional aquifers for six large streams in the upper Coastal Plain of South Carolina and parts of North Carolina and Georgia. Aquifers that sustain the base flow of both large and small streams are stratified into shallow and deep flow systems. Base-flow during dry conditions on main stems of large streams was assumed to be the discharge from the deep groundwater flow system. Six streams were analyzed: the Savannah, South and North Fork Edisto, Lynches, Pee Dee, and the Luber Rivers. Stream reaches in the Upper Coastal Plain were studied because of the relatively large aquifer discharge in these areas in comparison to the lower Coastal Plain. Estimates of discharge from the deep groundwater flow system to the six large streams averaged 1.8 cu ft/sec/mi of stream and 0.11 cu ft/sec/sq mi of surface drainage area. The estimates were made by subtracting all tributary inflows from the discharge gain between two gaging stations on a large stream during an extreme low-flow period. These estimates pertain only to flow in the deep groundwater flow system. Shallow flow systems and total base flow are > flow in the deep system. (USGS)

Automatic Classification of volcano-seismic events based on Deep Neural Networks.

NASA Astrophysics Data System (ADS)

Titos Luzón, M.; Bueno Rodriguez, A.; Garcia Martinez, L.; Benitez, C.; Ibáñez, J. M.

2017-12-01

Seismic monitoring of active volcanoes is a popular remote sensing technique to detect seismic activity, often associated to energy exchanges between the volcano and the environment. As a result, seismographs register a wide range of volcano-seismic signals that reflect the nature and underlying physics of volcanic processes. Machine learning and signal processing techniques provide an appropriate framework to analyze such data. In this research, we propose a new classification framework for seismic events based on deep neural networks. Deep neural networks are composed by multiple processing layers, and can discover intrinsic patterns from the data itself. Internal parameters can be initialized using a greedy unsupervised pre-training stage, leading to an efficient training of fully connected architectures. We aim to determine the robustness of these architectures as classifiers of seven different types of seismic events recorded at "Volcán de Fuego" (Colima, Mexico). Two deep neural networks with different pre-training strategies are studied: stacked denoising autoencoder and deep belief networks. Results are compared to existing machine learning algorithms (SVM, Random Forest, Multilayer Perceptron). We used 5 LPC coefficients over three non-overlapping segments as training features in order to characterize temporal evolution, avoid redundancy and encode the signal, regardless of its duration. Experimental results show that deep architectures can classify seismic events with higher accuracy than classical algorithms, attaining up to 92% recognition accuracy. Pre-training initialization helps these models to detect events that occur simultaneously in time (such explosions and rockfalls), increase robustness against noisy inputs, and provide better generalization. These results demonstrate deep neural networks are robust classifiers, and can be deployed in real-environments to monitor the seismicity of restless volcanoes.

Evolving Deep Networks Using HPC

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

Young, Steven R.; Rose, Derek C.; Johnston, Travis

While a large number of deep learning networks have been studied and published that produce outstanding results on natural image datasets, these datasets only make up a fraction of those to which deep learning can be applied. These datasets include text data, audio data, and arrays of sensors that have very different characteristics than natural images. As these “best” networks for natural images have been largely discovered through experimentation and cannot be proven optimal on some theoretical basis, there is no reason to believe that they are the optimal network for these drastically different datasets. Hyperparameter search is thus oftenmore » a very important process when applying deep learning to a new problem. In this work we present an evolutionary approach to searching the possible space of network hyperparameters and construction that can scale to 18, 000 nodes. This approach is applied to datasets of varying types and characteristics where we demonstrate the ability to rapidly find best hyperparameters in order to enable practitioners to quickly iterate between idea and result.« less

Multispectral embedding-based deep neural network for three-dimensional human pose recovery

NASA Astrophysics Data System (ADS)

Yu, Jialin; Sun, Jifeng

2018-01-01

Monocular image-based three-dimensional (3-D) human pose recovery aims to retrieve 3-D poses using the corresponding two-dimensional image features. Therefore, the pose recovery performance highly depends on the image representations. We propose a multispectral embedding-based deep neural network (MSEDNN) to automatically obtain the most discriminative features from multiple deep convolutional neural networks and then embed their penultimate fully connected layers into a low-dimensional manifold. This compact manifold can explore not only the optimum output from multiple deep networks but also the complementary properties of them. Furthermore, the distribution of each hierarchy discriminative manifold is sufficiently smooth so that the training process of our MSEDNN can be effectively implemented only using few labeled data. Our proposed network contains a body joint detector and a human pose regressor that are jointly trained. Extensive experiments conducted on four databases show that our proposed MSEDNN can achieve the best recovery performance compared with the state-of-the-art methods.

Sentiment analysis: a comparison of deep learning neural network algorithm with SVM and naϊve Bayes for Indonesian text

NASA Astrophysics Data System (ADS)

Calvin Frans Mariel, Wahyu; Mariyah, Siti; Pramana, Setia

2018-03-01

Deep learning is a new era of machine learning techniques that essentially imitate the structure and function of the human brain. It is a development of deeper Artificial Neural Network (ANN) that uses more than one hidden layer. Deep Learning Neural Network has a great ability on recognizing patterns from various data types such as picture, audio, text, and many more. In this paper, the authors tries to measure that algorithm’s ability by applying it into the text classification. The classification task herein is done by considering the content of sentiment in a text which is also called as sentiment analysis. By using several combinations of text preprocessing and feature extraction techniques, we aim to compare the precise modelling results of Deep Learning Neural Network with the other two commonly used algorithms, the Naϊve Bayes and Support Vector Machine (SVM). This algorithm comparison uses Indonesian text data with balanced and unbalanced sentiment composition. Based on the experimental simulation, Deep Learning Neural Network clearly outperforms the Naϊve Bayes and SVM and offers a better F-1 Score while for the best feature extraction technique which improves that modelling result is Bigram.

Reprint of: The evil of sluits: a re-assessment of soil erosion in the Karoo of South Africa as portrayed in century-old sources.

PubMed

Rowntree, K M

2014-06-01

Deep, linear gullies are a common feature of the present landscape of the Karoo of South Africa, where they were known locally in the early twentieth century as 'sluits'. Recent research has shown that many of these features are now stable and are no longer significant sediment sources, although they are efficient connectors in the landscape. Because most of the gully networks predate the first aerial photographs, little is known in the scientific literature about the timing of their formation. One secondary source, however, throws interesting light on the origin of these features, and the early response by landowners to their rehabilitation. The Agricultural Journal of the Cape of Good Hope at the turn of the Twentieth Century carried a number of articles by farmers and agricultural officers concerning the "evil of sluits". The authors gave accounts of widespread incision of valley bottoms by deep, wide gullies. Many of these gullies had been in existence for some thirty years but apparently had formed within living memory. A number of attempts to prevent further erosion had been put in place at the time of writing. This paper presents a review of land degradation, specifically gully erosion, and rehabilitation recommendations as given by authors writing in this journal. It reflects on the findings in the context of assessing land degradation processes through the local knowledge portrayed in the journal. Copyright © 2014. Published by Elsevier Ltd.

The evil of sluits: a re-assessment of soil erosion in the Karoo of South Africa as portrayed in century-old sources.

PubMed

Rowntree, K M

2013-11-30

Deep, linear gullies are a common feature of the present landscape of the Karoo of South Africa, where they were known locally in the early twentieth century as 'sluits'. Recent research has shown that many of these features are now stable and are no longer significant sediment sources, although they are efficient connectors in the landscape. Because most of the gully networks predate the first aerial photographs, little is known in the scientific literature about the timing of their formation. One secondary source, however, throws interesting light on the origin of these features, and the early response by landowners to their rehabilitation. The Agricultural Journal of the Cape of Good Hope at the turn of the Twentieth Century carried a number of articles by farmers and agricultural officers concerning the "evil of sluits". The authors gave accounts of widespread incision of valley bottoms by deep, wide gullies. Many of these gullies had been in existence for some thirty years but apparently had formed within living memory. A number of attempts to prevent further erosion had been put in place at the time of writing. This paper presents a review of land degradation, specifically gully erosion, and rehabilitation recommendations as given by authors writing in this journal. It reflects on the findings in the context of assessing land degradation processes through the local knowledge portrayed in the journal. Copyright © 2013 Elsevier Ltd. All rights reserved.

Sub-microradian pointing for deep space optical telecommunications network

NASA Technical Reports Server (NTRS)

Ortiz, G.; Lee, S.; Alexander, J.

2001-01-01

This presentation will cover innovative hardware, algorithms, architectures, techniques and recent laboratory results that are applicable to all deep space optical communication links, such as the Mars Telecommunication Network to future interstellar missions.

Deep Learning: A Primer for Radiologists.

PubMed

Chartrand, Gabriel; Cheng, Phillip M; Vorontsov, Eugene; Drozdzal, Michal; Turcotte, Simon; Pal, Christopher J; Kadoury, Samuel; Tang, An

2017-01-01

Deep learning is a class of machine learning methods that are gaining success and attracting interest in many domains, including computer vision, speech recognition, natural language processing, and playing games. Deep learning methods produce a mapping from raw inputs to desired outputs (eg, image classes). Unlike traditional machine learning methods, which require hand-engineered feature extraction from inputs, deep learning methods learn these features directly from data. With the advent of large datasets and increased computing power, these methods can produce models with exceptional performance. These models are multilayer artificial neural networks, loosely inspired by biologic neural systems. Weighted connections between nodes (neurons) in the network are iteratively adjusted based on example pairs of inputs and target outputs by back-propagating a corrective error signal through the network. For computer vision tasks, convolutional neural networks (CNNs) have proven to be effective. Recently, several clinical applications of CNNs have been proposed and studied in radiology for classification, detection, and segmentation tasks. This article reviews the key concepts of deep learning for clinical radiologists, discusses technical requirements, describes emerging applications in clinical radiology, and outlines limitations and future directions in this field. Radiologists should become familiar with the principles and potential applications of deep learning in medical imaging. © RSNA, 2017.

Using deep learning in image hyper spectral segmentation, classification, and detection

NASA Astrophysics Data System (ADS)

Zhao, Xiuying; Su, Zhenyu

2018-02-01

Recent years have shown that deep learning neural networks are a valuable tool in the field of computer vision. Deep learning method can be used in applications like remote sensing such as Land cover Classification, Detection of Vehicle in Satellite Images, Hyper spectral Image classification. This paper addresses the use of the deep learning artificial neural network in Satellite image segmentation. Image segmentation plays an important role in image processing. The hue of the remote sensing image often has a large hue difference, which will result in the poor display of the images in the VR environment. Image segmentation is a pre processing technique applied to the original images and splits the image into many parts which have different hue to unify the color. Several computational models based on supervised, unsupervised, parametric, probabilistic region based image segmentation techniques have been proposed. Recently, one of the machine learning technique known as, deep learning with convolution neural network has been widely used for development of efficient and automatic image segmentation models. In this paper, we focus on study of deep neural convolution network and its variants for automatic image segmentation rather than traditional image segmentation strategies.

January 30, 1997 eruptive event on Kilauea Volcano, Hawaii, as monitored by continuous GPS

USGS Publications Warehouse

Owen, S.; Segall, P.; Lisowski, M.; Miklius, Asta; Murray, M.; Bevis, M.; Foster, J.

2000-01-01

A continuous Global Positioning System (GPS) network on Kilauea Volcano captured the most recent fissure eruption in Kilauea's East Rift Zone (ERZ) in unprecedented spatial and temporal detail. The short eruption drained the lava pond at Pu'u O' o, leading to a two month long pause in its on-going eruption. Models of the GPS data indicate that the intrusion's bottom edge extended to only 2.4 km. Continuous GPS data reveal rift opening 8 hours prior to the eruption. Absence of precursory summit inflation rules out magma storage overpressurization as the eruption's cause. We infer that stresses in the shallow rift created by the continued deep rift dilation and slip on the south flank decollement caused the rift intrusion.

ShapeShop: Towards Understanding Deep Learning Representations via Interactive Experimentation.

PubMed

Hohman, Fred; Hodas, Nathan; Chau, Duen Horng

2017-05-01

Deep learning is the driving force behind many recent technologies; however, deep neural networks are often viewed as "black-boxes" due to their internal complexity that is hard to understand. Little research focuses on helping people explore and understand the relationship between a user's data and the learned representations in deep learning models. We present our ongoing work, ShapeShop, an interactive system for visualizing and understanding what semantics a neural network model has learned. Built using standard web technologies, ShapeShop allows users to experiment with and compare deep learning models to help explore the robustness of image classifiers.

A study of the relationship between food group recommendations and perceived stress: findings from black women in the Deep South.

PubMed

Carson, Tiffany L; Desmond, Renee; Hardy, Sharonda; Townsend, Sh'Nese; Ard, Jamy D; Meneses, Karen; Partridge, Edward E; Baskin, Monica L

2015-01-01

Black women in the Deep South experience excess morbidity/mortality from obesity-related diseases, which may be partially attributable to poor diet. One reason for poor dietary intake may be high stress, which has been associated with unhealthy diets in other groups. Limited data are available regarding dietary patterns of black women in the Deep South and to our knowledge no studies have been published exploring relationships between stress and dietary patterns among this group. This cross-sectional study explored the relationship between stress and adherence to food group recommendations among black women in the Deep South. Participants (n = 355) provided demographic, anthropometric, stress (PSS-10), and dietary (NCI ASA-24 hour recall) data. Participants were obese (BMI = 36.5 kg/m(2)) and reported moderate stress (PSS-10 score = 16) and minimal adherence to Dietary Guidelines for Americans food group recommendations (1/3 did not meet recommendations for any food group). Participants reporting higher stress had higher BMIs than those reporting lower stress. There was no observed relationship between stress and dietary intake in this sample. Based on these study findings, which are limited by potential misreporting of dietary intake and limited variability in stress measure outcomes, there is insufficient evidence to support a relationship between stress and dietary intake.

DeepNAT: Deep convolutional neural network for segmenting neuroanatomy.

PubMed

Wachinger, Christian; Reuter, Martin; Klein, Tassilo

2018-04-15

We introduce DeepNAT, a 3D Deep convolutional neural network for the automatic segmentation of NeuroAnaTomy in T1-weighted magnetic resonance images. DeepNAT is an end-to-end learning-based approach to brain segmentation that jointly learns an abstract feature representation and a multi-class classification. We propose a 3D patch-based approach, where we do not only predict the center voxel of the patch but also neighbors, which is formulated as multi-task learning. To address a class imbalance problem, we arrange two networks hierarchically, where the first one separates foreground from background, and the second one identifies 25 brain structures on the foreground. Since patches lack spatial context, we augment them with coordinates. To this end, we introduce a novel intrinsic parameterization of the brain volume, formed by eigenfunctions of the Laplace-Beltrami operator. As network architecture, we use three convolutional layers with pooling, batch normalization, and non-linearities, followed by fully connected layers with dropout. The final segmentation is inferred from the probabilistic output of the network with a 3D fully connected conditional random field, which ensures label agreement between close voxels. The roughly 2.7million parameters in the network are learned with stochastic gradient descent. Our results show that DeepNAT compares favorably to state-of-the-art methods. Finally, the purely learning-based method may have a high potential for the adaptation to young, old, or diseased brains by fine-tuning the pre-trained network with a small training sample on the target application, where the availability of larger datasets with manual annotations may boost the overall segmentation accuracy in the future. Copyright © 2017 Elsevier Inc. All rights reserved.

A Hybrid Spectral Clustering and Deep Neural Network Ensemble Algorithm for Intrusion Detection in Sensor Networks

PubMed Central

Ma, Tao; Wang, Fen; Cheng, Jianjun; Yu, Yang; Chen, Xiaoyun

2016-01-01

The development of intrusion detection systems (IDS) that are adapted to allow routers and network defence systems to detect malicious network traffic disguised as network protocols or normal access is a critical challenge. This paper proposes a novel approach called SCDNN, which combines spectral clustering (SC) and deep neural network (DNN) algorithms. First, the dataset is divided into k subsets based on sample similarity using cluster centres, as in SC. Next, the distance between data points in a testing set and the training set is measured based on similarity features and is fed into the deep neural network algorithm for intrusion detection. Six KDD-Cup99 and NSL-KDD datasets and a sensor network dataset were employed to test the performance of the model. These experimental results indicate that the SCDNN classifier not only performs better than backpropagation neural network (BPNN), support vector machine (SVM), random forest (RF) and Bayes tree models in detection accuracy and the types of abnormal attacks found. It also provides an effective tool of study and analysis of intrusion detection in large networks. PMID:27754380

A Hybrid Spectral Clustering and Deep Neural Network Ensemble Algorithm for Intrusion Detection in Sensor Networks.

PubMed

Ma, Tao; Wang, Fen; Cheng, Jianjun; Yu, Yang; Chen, Xiaoyun

2016-10-13

The development of intrusion detection systems (IDS) that are adapted to allow routers and network defence systems to detect malicious network traffic disguised as network protocols or normal access is a critical challenge. This paper proposes a novel approach called SCDNN, which combines spectral clustering (SC) and deep neural network (DNN) algorithms. First, the dataset is divided into k subsets based on sample similarity using cluster centres, as in SC. Next, the distance between data points in a testing set and the training set is measured based on similarity features and is fed into the deep neural network algorithm for intrusion detection. Six KDD-Cup99 and NSL-KDD datasets and a sensor network dataset were employed to test the performance of the model. These experimental results indicate that the SCDNN classifier not only performs better than backpropagation neural network (BPNN), support vector machine (SVM), random forest (RF) and Bayes tree models in detection accuracy and the types of abnormal attacks found. It also provides an effective tool of study and analysis of intrusion detection in large networks.

A Composite Model of Wound Segmentation Based on Traditional Methods and Deep Neural Networks

PubMed Central

Wang, Changjian; Liu, Xiaohui; Jin, Shiyao

2018-01-01

Wound segmentation plays an important supporting role in the wound observation and wound healing. Current methods of image segmentation include those based on traditional process of image and those based on deep neural networks. The traditional methods use the artificial image features to complete the task without large amounts of labeled data. Meanwhile, the methods based on deep neural networks can extract the image features effectively without the artificial design, but lots of training data are required. Combined with the advantages of them, this paper presents a composite model of wound segmentation. The model uses the skin with wound detection algorithm we designed in the paper to highlight image features. Then, the preprocessed images are segmented by deep neural networks. And semantic corrections are applied to the segmentation results at last. The model shows a good performance in our experiment. PMID:29955227

Machine Learning and Quantum Mechanics

NASA Astrophysics Data System (ADS)

Chapline, George

The author has previously pointed out some similarities between selforganizing neural networks and quantum mechanics. These types of neural networks were originally conceived of as away of emulating the cognitive capabilities of the human brain. Recently extensions of these networks, collectively referred to as deep learning networks, have strengthened the connection between self-organizing neural networks and human cognitive capabilities. In this note we consider whether hardware quantum devices might be useful for emulating neural networks with human-like cognitive capabilities, or alternatively whether implementations of deep learning neural networks using conventional computers might lead to better algorithms for solving the many body Schrodinger equation.

The Deep Space Network: An instrument for radio astronomy research

NASA Technical Reports Server (NTRS)

Renzetti, N. A.; Levy, G. S.; Kuiper, T. B. H.; Walken, P. R.; Chandlee, R. C.

1988-01-01

The NASA Deep Space Network operates and maintains the Earth-based two-way communications link for unmanned spacecraft exploring the solar system. It is NASA's policy to also make the Network's facilities available for radio astronomy observations. The Network's microwave communication systems and facilities are being continually upgraded. This revised document, first published in 1982, describes the Network's current radio astronomy capabilities and future capabilities that will be made available by the ongoing Network upgrade. The Bibliography, which includes published papers and articles resulting from radio astronomy observations conducted with Network facilities, has been updated to include papers to May 1987.

Deep space communication - A one billion mile noisy channel

NASA Technical Reports Server (NTRS)

Smith, J. G.

1982-01-01

Deep space exploration is concerned with the study of natural phenomena in the solar system with the aid of measurements made at spacecraft on deep space missions. Deep space communication refers to communication between earth and spacecraft in deep space. The Deep Space Network is an earth-based facility employed for deep space communication. It includes a network of large tracking antennas located at various positions around the earth. The goals and achievements of deep space exploration over the past 20 years are discussed along with the broad functional requirements of deep space missions. Attention is given to the differences in space loss between communication satellites and deep space vehicles, effects of the long round-trip light time on spacecraft autonomy, requirements for the use of massive nuclear power plants on spacecraft at large distances from the sun, and the kinds of scientific return provided by a deep space mission. Problems concerning a deep space link of one billion miles are also explored.

The Telecommunications and Data Acquisition Report. [Deep Space Network

NASA Technical Reports Server (NTRS)

Posner, E. C. (Editor)

1986-01-01

This publication, one of a series formerly titled The Deep Space Network Progress Report, documents DSN progress in flight project support, tracking and data acquisition research and technology, network engineering, hardware and software implementation, and operations. In addition, developments in Earth-based radio technology as applied to geodynamics, astrophysics and the radio search for extraterrestrial intelligence are reported.

A Study of Complex Deep Learning Networks on High Performance, Neuromorphic, and Quantum Computers

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

Potok, Thomas E; Schuman, Catherine D; Young, Steven R

Current Deep Learning models use highly optimized convolutional neural networks (CNN) trained on large graphical processing units (GPU)-based computers with a fairly simple layered network topology, i.e., highly connected layers, without intra-layer connections. Complex topologies have been proposed, but are intractable to train on current systems. Building the topologies of the deep learning network requires hand tuning, and implementing the network in hardware is expensive in both cost and power. In this paper, we evaluate deep learning models using three different computing architectures to address these problems: quantum computing to train complex topologies, high performance computing (HPC) to automatically determinemore » network topology, and neuromorphic computing for a low-power hardware implementation. Due to input size limitations of current quantum computers we use the MNIST dataset for our evaluation. The results show the possibility of using the three architectures in tandem to explore complex deep learning networks that are untrainable using a von Neumann architecture. We show that a quantum computer can find high quality values of intra-layer connections and weights, while yielding a tractable time result as the complexity of the network increases; a high performance computer can find optimal layer-based topologies; and a neuromorphic computer can represent the complex topology and weights derived from the other architectures in low power memristive hardware. This represents a new capability that is not feasible with current von Neumann architecture. It potentially enables the ability to solve very complicated problems unsolvable with current computing technologies.« less

Deep neural networks for direct, featureless learning through observation: The case of two-dimensional spin models

NASA Astrophysics Data System (ADS)

Mills, Kyle; Tamblyn, Isaac

2018-03-01

We demonstrate the capability of a convolutional deep neural network in predicting the nearest-neighbor energy of the 4 ×4 Ising model. Using its success at this task, we motivate the study of the larger 8 ×8 Ising model, showing that the deep neural network can learn the nearest-neighbor Ising Hamiltonian after only seeing a vanishingly small fraction of configuration space. Additionally, we show that the neural network has learned both the energy and magnetization operators with sufficient accuracy to replicate the low-temperature Ising phase transition. We then demonstrate the ability of the neural network to learn other spin models, teaching the convolutional deep neural network to accurately predict the long-range interaction of a screened Coulomb Hamiltonian, a sinusoidally attenuated screened Coulomb Hamiltonian, and a modified Potts model Hamiltonian. In the case of the long-range interaction, we demonstrate the ability of the neural network to recover the phase transition with equivalent accuracy to the numerically exact method. Furthermore, in the case of the long-range interaction, the benefits of the neural network become apparent; it is able to make predictions with a high degree of accuracy, and do so 1600 times faster than a CUDA-optimized exact calculation. Additionally, we demonstrate how the neural network succeeds at these tasks by looking at the weights learned in a simplified demonstration.

Deep RNNs for video denoising

NASA Astrophysics Data System (ADS)

Chen, Xinyuan; Song, Li; Yang, Xiaokang

2016-09-01

Video denoising can be described as the problem of mapping from a specific length of noisy frames to clean one. We propose a deep architecture based on Recurrent Neural Network (RNN) for video denoising. The model learns a patch-based end-to-end mapping between the clean and noisy video sequences. It takes the corrupted video sequences as the input and outputs the clean one. Our deep network, which we refer to as deep Recurrent Neural Networks (deep RNNs or DRNNs), stacks RNN layers where each layer receives the hidden state of the previous layer as input. Experiment shows (i) the recurrent architecture through temporal domain extracts motion information and does favor to video denoising, and (ii) deep architecture have large enough capacity for expressing mapping relation between corrupted videos as input and clean videos as output, furthermore, (iii) the model has generality to learned different mappings from videos corrupted by different types of noise (e.g., Poisson-Gaussian noise). By training on large video databases, we are able to compete with some existing video denoising methods.

Deep learning for medical image segmentation - using the IBM TrueNorth neurosynaptic system

NASA Astrophysics Data System (ADS)

Moran, Steven; Gaonkar, Bilwaj; Whitehead, William; Wolk, Aidan; Macyszyn, Luke; Iyer, Subramanian S.

2018-03-01

Deep convolutional neural networks have found success in semantic image segmentation tasks in computer vision and medical imaging. These algorithms are executed on conventional von Neumann processor architectures or GPUs. This is suboptimal. Neuromorphic processors that replicate the structure of the brain are better-suited to train and execute deep learning models for image segmentation by relying on massively-parallel processing. However, given that they closely emulate the human brain, on-chip hardware and digital memory limitations also constrain them. Adapting deep learning models to execute image segmentation tasks on such chips, requires specialized training and validation. In this work, we demonstrate for the first-time, spinal image segmentation performed using a deep learning network implemented on neuromorphic hardware of the IBM TrueNorth Neurosynaptic System and validate the performance of our network by comparing it to human-generated segmentations of spinal vertebrae and disks. To achieve this on neuromorphic hardware, the training model constrains the coefficients of individual neurons to {-1,0,1} using the Energy Efficient Deep Neuromorphic (EEDN)1 networks training algorithm. Given the 1 million neurons and 256 million synapses, the scale and size of the neural network implemented by the IBM TrueNorth allows us to execute the requisite mapping between segmented images and non-uniform intensity MR images >20 times faster than on a GPU-accelerated network and using <0.1 W. This speed and efficiency implies that a trained neuromorphic chip can be deployed in intra-operative environments where real-time medical image segmentation is necessary.

Fiber Orientation Estimation Guided by a Deep Network.

PubMed

Ye, Chuyang; Prince, Jerry L

2017-09-01

Diffusion magnetic resonance imaging (dMRI) is currently the only tool for noninvasively imaging the brain's white matter tracts. The fiber orientation (FO) is a key feature computed from dMRI for tract reconstruction. Because the number of FOs in a voxel is usually small, dictionary-based sparse reconstruction has been used to estimate FOs. However, accurate estimation of complex FO configurations in the presence of noise can still be challenging. In this work we explore the use of a deep network for FO estimation in a dictionary-based framework and propose an algorithm named Fiber Orientation Reconstruction guided by a Deep Network (FORDN). FORDN consists of two steps. First, we use a smaller dictionary encoding coarse basis FOs to represent diffusion signals. To estimate the mixture fractions of the dictionary atoms, a deep network is designed to solve the sparse reconstruction problem. Second, the coarse FOs inform the final FO estimation, where a larger dictionary encoding a dense basis of FOs is used and a weighted ℓ 1 -norm regularized least squares problem is solved to encourage FOs that are consistent with the network output. FORDN was evaluated and compared with state-of-the-art algorithms that estimate FOs using sparse reconstruction on simulated and typical clinical dMRI data. The results demonstrate the benefit of using a deep network for FO estimation.

Iterative deep convolutional encoder-decoder network for medical image segmentation.

PubMed

Jung Uk Kim; Hak Gu Kim; Yong Man Ro

2017-07-01

In this paper, we propose a novel medical image segmentation using iterative deep learning framework. We have combined an iterative learning approach and an encoder-decoder network to improve segmentation results, which enables to precisely localize the regions of interest (ROIs) including complex shapes or detailed textures of medical images in an iterative manner. The proposed iterative deep convolutional encoder-decoder network consists of two main paths: convolutional encoder path and convolutional decoder path with iterative learning. Experimental results show that the proposed iterative deep learning framework is able to yield excellent medical image segmentation performances for various medical images. The effectiveness of the proposed method has been proved by comparing with other state-of-the-art medical image segmentation methods.

Deep space network support of the manned space flight network for Apollo, volume 3. [support for Apollo 14, 15, 16, and 17 flights

NASA Technical Reports Server (NTRS)

Hartley, R. B.

1974-01-01

The Deep Space Network (DSN) activities in support of Project Apollo during the period of 1971 and 1972 are reported. Beginning with the Apollo 14 mission and concluding with the Apollo 17 mission, the narrative includes, (1) a mission description, (2) the NASA support requirements placed on the DSN, and, (3) a comprehensive account of the support activities provided by each committed DSN deep space communication station. Associated equipment and activities of the three elements of the DSN (the Deep Space Instrumentation Facility (DSIF), the Space Flight Operations Facility (SFOF), and the Ground Communications Facility (GCF)) used in meeting the radio-metric and telemetry demands of the missions are documented.

Deep convolutional neural networks for automatic classification of gastric carcinoma using whole slide images in digital histopathology.

PubMed

Sharma, Harshita; Zerbe, Norman; Klempert, Iris; Hellwich, Olaf; Hufnagl, Peter

2017-11-01

Deep learning using convolutional neural networks is an actively emerging field in histological image analysis. This study explores deep learning methods for computer-aided classification in H&E stained histopathological whole slide images of gastric carcinoma. An introductory convolutional neural network architecture is proposed for two computerized applications, namely, cancer classification based on immunohistochemical response and necrosis detection based on the existence of tumor necrosis in the tissue. Classification performance of the developed deep learning approach is quantitatively compared with traditional image analysis methods in digital histopathology requiring prior computation of handcrafted features, such as statistical measures using gray level co-occurrence matrix, Gabor filter-bank responses, LBP histograms, gray histograms, HSV histograms and RGB histograms, followed by random forest machine learning. Additionally, the widely known AlexNet deep convolutional framework is comparatively analyzed for the corresponding classification problems. The proposed convolutional neural network architecture reports favorable results, with an overall classification accuracy of 0.6990 for cancer classification and 0.8144 for necrosis detection. Copyright © 2017 Elsevier Ltd. All rights reserved.

The Deep Space Network

NASA Technical Reports Server (NTRS)

1975-01-01

Work accomplished on the Deep Space Network (DSN) was described, including the following topics: supporting research and technology, advanced development and engineering, system implementation, and DSN operations pertaining to mission-independent or multiple-mission development as well as to support of flight projects.

Assessing the Linguistic Productivity of Unsupervised Deep Neural Networks

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

Phillips, Lawrence A.; Hodas, Nathan O.

Increasingly, cognitive scientists have demonstrated interest in applying tools from deep learning. One use for deep learning is in language acquisition where it is useful to know if a linguistic phenomenon can be learned through domain-general means. To assess whether unsupervised deep learning is appropriate, we first pose a smaller question: Can unsupervised neural networks apply linguistic rules productively, using them in novel situations. We draw from the literature on determiner/noun productivity by training an unsupervised, autoencoder network measuring its ability to combine nouns with determiners. Our simple autoencoder creates combinations it has not previously encountered, displaying a degree ofmore » overlap similar to actual children. While this preliminary work does not provide conclusive evidence for productivity, it warrants further investigation with more complex models. Further, this work helps lay the foundations for future collaboration between the deep learning and cognitive science communities.« less

A statewide biomedical communications network for South Carolina.

PubMed

Mangiaracina, J; Sawyer, W A

1976-03-01

In 1972, the Medical University of South Carolina was awarded a contract to establish 4 Area Health Education Centers in South Carolina. These centers, based in community hospitals, provide residency programs, clinical instruction for students, and continuing education programs for health professionals. In late 1974, contractual agreements between the Medical University of South Carolina's Library/Learning Resource Center and the Area Health Education Centers were negotiated to provide book and nonbook learning materials to all health practitioners in South Carolina. The history and the functions of the resulting network and evaluation of audiovisual and self-instructional learning materials procured and distributed by the network are described.

Deep neural network for traffic sign recognition systems: An analysis of spatial transformers and stochastic optimisation methods.

PubMed

Arcos-García, Álvaro; Álvarez-García, Juan A; Soria-Morillo, Luis M

2018-03-01

This paper presents a Deep Learning approach for traffic sign recognition systems. Several classification experiments are conducted over publicly available traffic sign datasets from Germany and Belgium using a Deep Neural Network which comprises Convolutional layers and Spatial Transformer Networks. Such trials are built to measure the impact of diverse factors with the end goal of designing a Convolutional Neural Network that can improve the state-of-the-art of traffic sign classification task. First, different adaptive and non-adaptive stochastic gradient descent optimisation algorithms such as SGD, SGD-Nesterov, RMSprop and Adam are evaluated. Subsequently, multiple combinations of Spatial Transformer Networks placed at distinct positions within the main neural network are analysed. The recognition rate of the proposed Convolutional Neural Network reports an accuracy of 99.71% in the German Traffic Sign Recognition Benchmark, outperforming previous state-of-the-art methods and also being more efficient in terms of memory requirements. Copyright © 2018 Elsevier Ltd. All rights reserved.

Break-up of the Atlantic deep western boundary current into eddies at 8 degrees S.

PubMed

Dengler, M; Schott, F A; Eden, C; Brandt, P; Fischer, J; Zantopp, R J

2004-12-23

The existence in the ocean of deep western boundary currents, which connect the high-latitude regions where deep water is formed with upwelling regions as part of the global ocean circulation, was postulated more than 40 years ago. These ocean currents have been found adjacent to the continental slopes of all ocean basins, and have core depths between 1,500 and 4,000 m. In the Atlantic Ocean, the deep western boundary current is estimated to carry (10-40) x 10(6) m3 s(-1) of water, transporting North Atlantic Deep Water--from the overflow regions between Greenland and Scotland and from the Labrador Sea--into the South Atlantic and the Antarctic circumpolar current. Here we present direct velocity and water mass observations obtained in the period 2000 to 2003, as well as results from a numerical ocean circulation model, showing that the Atlantic deep western boundary current breaks up at 8 degrees S. Southward of this latitude, the transport of North Atlantic Deep Water into the South Atlantic Ocean is accomplished by migrating eddies, rather than by a continuous flow. Our model simulation indicates that the deep western boundary current breaks up into eddies at the present intensity of meridional overturning circulation. For weaker overturning, continuation as a stable, laminar boundary flow seems possible.

Simple techniques for improving deep neural network outcomes on commodity hardware

NASA Astrophysics Data System (ADS)

Colina, Nicholas Christopher A.; Perez, Carlos E.; Paraan, Francis N. C.

2017-08-01

We benchmark improvements in the performance of deep neural networks (DNN) on the MNIST data test upon imple-menting two simple modifications to the algorithm that have little overhead computational cost. First is GPU parallelization on a commodity graphics card, and second is initializing the DNN with random orthogonal weight matrices prior to optimization. Eigenspectra analysis of the weight matrices reveal that the initially orthogonal matrices remain nearly orthogonal after training. The probability distributions from which these orthogonal matrices are drawn are also shown to significantly affect the performance of these deep neural networks.

Deep structure of Llaima Volcano from seismic ambient noise tomography: Preliminary results

NASA Astrophysics Data System (ADS)

Franco, L.; Mikesell, T. D.; Rodd, R.; Lees, J. M.; Johnson, J. B.; Ronan, T.

2015-12-01

The ambient seismic noise tomography (ANT) method has become an important tool to image crustal structures and magmatic bodies at volcanoes. The frequency band of ambient noise provides complimentary data and added resolution to the deeper volcanic structures when compared to traditional tomography based on local earthquakes. The Llaima Volcano (38° 41.9' S and 71° 43.8' W) is a stratovolcano of basaltic-andesitic composition. Llaima is located in the South Volcanic Zone (ZVS) of the Andes and is listed as one of the most active volcanoes in South America, with a long documented historical record dating back to 1640. Llaima experienced violent eruptions in 1927 and 1957 (Naranjo and Moreno, 1991), and its last eruptive cycle (2008-2010) is considered the most important after the 1957 eruption. Lacking seismic constraints on the deep structure under Llaima, petrologic data have suggested the presence of magmatic bodies (dikes). These bodies likely play an important role in the eruptive dynamics of Llaima (Bouvet de Maisonneuve, C., et al 2012). Analysis of the 2008-2010 seismicity shows a southern zone (approx. 15 km from the Llaima summit) where there were many Very Long Period events occurring prior to the eruptions. This is in agreement with a deformation zone determined by InSAR analysis (Fournier et al, 2010 and Bathke, 2011), but no geologic model based on geophysical imaging has been created yet. Beginning in 2009, staff from the Chilean Geological Survey (SERNAGEOMIN) started to install a permanent seismic network consisting of nine stations. These nine stations have allowed Chilean seismologists to closely monitor the activity at Llaima, but prevented a high-resolution tomographic imaging study. During the summer of 2015, a temporary seismic network consisting of 26 stations was installed around Llaima. In the work presented here, we analyze continuous waveforms recorded between January and April 2015 from a total of 35 broadband stations (permanent and temporary). This network covers the total area of Llaima and provides the first study aimed at revealing the volcanic structure of Llaima. Moreover this is one of the first attempts to perform high resolution ANT at a Chilean volcano. We will present our tomography results and our first geologic interpretations of Llaima volcanic structure.

Imaging Subsurface Velocity Structure Under the Borborema Province, NE Brazil, With Passive-Source Seismology: From Crust to Lithosphere and Beyond

NASA Astrophysics Data System (ADS)

Julia, J.; Nascimento, R.; Bastow, I. D.; Dias, R. C.; Pinheiro, A. G.; Farias do Nascimento, A.; Ferreira, J. M.; Fuck, R. A.

2013-05-01

The Borborema Province of NE Brazil can be regarded as a collage ofseveral terranes of Precambrian age that amalgamated during the Brasiliano-Pan African orogeny around 600 Ma. It comprises the northeasternmost corner of the South American continent and it is bounded by the São Francisco craton to the South, the Paleozoic Parnaiba basin to the West and a number of Mesozoic marginal basins to the North and East. The Cenozoic evolution of the Province is marked by the uplift of the Borborema Plateau and the coeval magmatism along two mutually orthogonal alignments: Macau-Queimadas, onshore and trending in the NS direction, and Fernando de Noronha-Mecejana, offshore and trending EW. Constraints on the geodynamical evolution of the Province come mostly from geochronological data and neotectonic markers, which have related this Cenozoic volcanism and the coeval plateau uplift to a small-scale convection cell that might have developed at the edge of the continent. Available seismic constraints on deep crustal and upper mantle structure to validate this interpretation, however, are scarce. In order to develop seismic constraints on deep crustal and upper mantle structure, a network of 16 short-period stations was deployed in 2011 under the Instituto Nacional de Ciência e Tecnologia de Estudos Tectônicos (INCT-ET) of the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). The stations complement an existing network of 16 broadband stations used for seismic monitoring of the Brazilian northeast. The combined network has an aperture of ˜400 km in the NE direction, ˜600 km in the NS direction, and an average inter-station spacing of ˜100 km and will operate for about 2 years. Tomographic images based on fundamental model surface-waves dispersion as well as ambient-noise cross-correlations and P- and S-wave travel-times are now being developed, along with detailed crustal-velocity models from the joint inversion of receiver functions and surface-wave dispersion, anisotropy constraints from SKS-splitting, and transition zone discontinuity topography from receiver function stacks. We expect that the new results will help shed light on the origin of the Cenozoic volcanism and uplift mechanism for the Borborema Province.

ShapeShop: Towards Understanding Deep Learning Representations via Interactive Experimentation

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

Hohman, Frederick M.; Hodas, Nathan O.; Chau, Duen Horng

Deep learning is the driving force behind many recent technologies; however, deep neural networks are often viewed as “black-boxes” due to their internal complexity that is hard to understand. Little research focuses on helping people explore and understand the relationship between a user’s data and the learned representations in deep learning models. We present our ongoing work, ShapeShop, an interactive system for visualizing and understanding what semantics a neural network model has learned. Built using standard web technologies, ShapeShop allows users to experiment with and compare deep learning models to help explore the robustness of image classifiers.

Age and Gender Differences in Social Network Composition and Social Support Among Older Rural South Africans: Findings From the HAALSI Study.

PubMed

Harling, Guy; Morris, Katherine Ann; Manderson, Lenore; Perkins, Jessica M; Berkman, Lisa F

2018-03-26

Drawing on the "Health and Aging in Africa: A Longitudinal Study of an INDEPTH community in South Africa" (HAALSI) baseline survey, we present data on older adults' social networks and receipt of social support in rural South Africa. We examine how age and gender differences in social network characteristics matched with patterns predicted by theories of choice- and constraint-based network contraction in older adults. We used regression analysis on data for 5,059 South African adults aged 40 and older. Older respondents reported fewer important social contacts and less frequent communication than their middle-aged peers, largely due to fewer nonkin connections. Network size difference between older and younger respondents was greater for women than for men. These gender and age differences were explicable by much higher levels of widowhood among older women compared to younger women and older men. There was no evidence for employment-related network contraction or selective retention of emotionally supportive ties. Marriage-related structural constraints impacted on older women's social networks in rural South Africa, but did not explain choice-based network contraction. These findings suggest that many older women in rural Africa, a growing population, may have an unmet need for social support.

A novel wavelet sequence based on deep bidirectional LSTM network model for ECG signal classification.

PubMed

Yildirim, Özal

2018-05-01

Long-short term memory networks (LSTMs), which have recently emerged in sequential data analysis, are the most widely used type of recurrent neural networks (RNNs) architecture. Progress on the topic of deep learning includes successful adaptations of deep versions of these architectures. In this study, a new model for deep bidirectional LSTM network-based wavelet sequences called DBLSTM-WS was proposed for classifying electrocardiogram (ECG) signals. For this purpose, a new wavelet-based layer is implemented to generate ECG signal sequences. The ECG signals were decomposed into frequency sub-bands at different scales in this layer. These sub-bands are used as sequences for the input of LSTM networks. New network models that include unidirectional (ULSTM) and bidirectional (BLSTM) structures are designed for performance comparisons. Experimental studies have been performed for five different types of heartbeats obtained from the MIT-BIH arrhythmia database. These five types are Normal Sinus Rhythm (NSR), Ventricular Premature Contraction (VPC), Paced Beat (PB), Left Bundle Branch Block (LBBB), and Right Bundle Branch Block (RBBB). The results show that the DBLSTM-WS model gives a high recognition performance of 99.39%. It has been observed that the wavelet-based layer proposed in the study significantly improves the recognition performance of conventional networks. This proposed network structure is an important approach that can be applied to similar signal processing problems. Copyright © 2018 Elsevier Ltd. All rights reserved.

Dissolved inorganic carbon isotopic composition of the Gulf of Mexico deep-water masses.

NASA Astrophysics Data System (ADS)

Quintanilla-Terminel, J. G.; Herguera, J. C.; Ferreira-Bartrina, V.; Hernández-Ayón, J. M.; Camacho-Ibar, V.

2014-12-01

This study provides new data for the establishment of a carbon biogeochemical dynamics baseline in the deep Gulf of Mexico (GM) based on carbon isotopes in dissolved inorganic carbon. Water samples from 40 deep-water stations south of 25˚N were collected during XIXIMI-2 cruise, July 2011, aboard BO/Justo Sierra. Vertical profiles of temperature, salinity and dissolved oxygen (DO) were further measured in each station. In the Stable Isotopes Laboratory at CICESE we determined the carbon isotopic composition of the dissolved inorganic carbon (DIC) (δ13CDIC). Remarkably, density, DO and δ13CCID profiles showed a clear difference between the Loop current and the deep-waters of the GM south of 25˚N. We found the following average δ13CCID values in the Loop current and in the deep-waters of the Gulf: subtropical underwater (SUW): 0.73±0.06‰ and 0.86±0.04‰; 18 degree water (18W): 0.76 ± 0.08‰ and 0.58± 0.06‰; North Atlantic central water (NACW): 0.77 ± 0.05‰ and 0.71 ± 0.09‰; South Atlantic central water (SACW): 0.80 ± 0.08‰ and 0.77 ± 0.07‰; Antartic intermediate water (AAIW): 1.00 ± 0.06‰ and 0.90 ± 0.08‰; North Atlantic deep water (NADW): 1.03 ± 0.06‰ and 1.01 ± 0.10‰. We will discuss how the biological component, δ13CCID-BIO, of subsurface water masses match very closely the apparent oxygen utilization relation described by Kroopnick, 1985, with the exception of SUW, and as a consequence the 18W is probably the water mass most affected by organic carbon remineralization processes in the GM south of 25˚N. We further show how these waters seem to store a larger proportion of anthropogenic carbon than the deeper water masses.

Vertical migration of municipal wastewater in deep injection well systems, South Florida, USA

NASA Astrophysics Data System (ADS)

Maliva, Robert G.; Guo, Weixing; Missimer, Thomas

2007-11-01

Deep well injection is widely used in South Florida, USA for wastewater disposal largely because of the presence of an injection zone (“boulder zone” of Floridan Aquifer System) that is capable of accepting very large quantities of fluids, in some wells over 75,000 m3/day. The greatest potential risk to public health associated with deep injection wells in South Florida is vertical migration of wastewater, containing pathogenic microorganisms and pollutants, into brackish-water aquifer zones that are being used for alternative water-supply projects such as aquifer storage and recovery. Upwards migration of municipal wastewater has occurred in a minority of South Florida injection systems. The results of solute-transport modeling using the SEAWAT program indicate that the measured vertical hydraulic conductivities of the rock matrix would allow for only minimal vertical migration. Fracturing at some sites increased the equivalent average vertical hydraulic conductivity of confining zone strata by approximately four orders of magnitude and allowed for vertical migration rates of up 80 m/year. Even where vertical migration was rapid, the documented transit times are likely long enough for the inactivation of pathogenic microorganisms.

Clinical Named Entity Recognition Using Deep Learning Models.

PubMed

Wu, Yonghui; Jiang, Min; Xu, Jun; Zhi, Degui; Xu, Hua

2017-01-01

Clinical Named Entity Recognition (NER) is a critical natural language processing (NLP) task to extract important concepts (named entities) from clinical narratives. Researchers have extensively investigated machine learning models for clinical NER. Recently, there have been increasing efforts to apply deep learning models to improve the performance of current clinical NER systems. This study examined two popular deep learning architectures, the Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN), to extract concepts from clinical texts. We compared the two deep neural network architectures with three baseline Conditional Random Fields (CRFs) models and two state-of-the-art clinical NER systems using the i2b2 2010 clinical concept extraction corpus. The evaluation results showed that the RNN model trained with the word embeddings achieved a new state-of-the- art performance (a strict F1 score of 85.94%) for the defined clinical NER task, outperforming the best-reported system that used both manually defined and unsupervised learning features. This study demonstrates the advantage of using deep neural network architectures for clinical concept extraction, including distributed feature representation, automatic feature learning, and long-term dependencies capture. This is one of the first studies to compare the two widely used deep learning models and demonstrate the superior performance of the RNN model for clinical NER.

MUFOLD-SS: New deep inception-inside-inception networks for protein secondary structure prediction.

PubMed

Fang, Chao; Shang, Yi; Xu, Dong

2018-05-01

Protein secondary structure prediction can provide important information for protein 3D structure prediction and protein functions. Deep learning offers a new opportunity to significantly improve prediction accuracy. In this article, a new deep neural network architecture, named the Deep inception-inside-inception (Deep3I) network, is proposed for protein secondary structure prediction and implemented as a software tool MUFOLD-SS. The input to MUFOLD-SS is a carefully designed feature matrix corresponding to the primary amino acid sequence of a protein, which consists of a rich set of information derived from individual amino acid, as well as the context of the protein sequence. Specifically, the feature matrix is a composition of physio-chemical properties of amino acids, PSI-BLAST profile, and HHBlits profile. MUFOLD-SS is composed of a sequence of nested inception modules and maps the input matrix to either eight states or three states of secondary structures. The architecture of MUFOLD-SS enables effective processing of local and global interactions between amino acids in making accurate prediction. In extensive experiments on multiple datasets, MUFOLD-SS outperformed the best existing methods and other deep neural networks significantly. MUFold-SS can be downloaded from http://dslsrv8.cs.missouri.edu/~cf797/MUFoldSS/download.html. © 2018 Wiley Periodicals, Inc.

Clinical Named Entity Recognition Using Deep Learning Models

PubMed Central

Wu, Yonghui; Jiang, Min; Xu, Jun; Zhi, Degui; Xu, Hua

2017-01-01

Clinical Named Entity Recognition (NER) is a critical natural language processing (NLP) task to extract important concepts (named entities) from clinical narratives. Researchers have extensively investigated machine learning models for clinical NER. Recently, there have been increasing efforts to apply deep learning models to improve the performance of current clinical NER systems. This study examined two popular deep learning architectures, the Convolutional Neural Network (CNN) and the Recurrent Neural Network (RNN), to extract concepts from clinical texts. We compared the two deep neural network architectures with three baseline Conditional Random Fields (CRFs) models and two state-of-the-art clinical NER systems using the i2b2 2010 clinical concept extraction corpus. The evaluation results showed that the RNN model trained with the word embeddings achieved a new state-of-the- art performance (a strict F1 score of 85.94%) for the defined clinical NER task, outperforming the best-reported system that used both manually defined and unsupervised learning features. This study demonstrates the advantage of using deep neural network architectures for clinical concept extraction, including distributed feature representation, automatic feature learning, and long-term dependencies capture. This is one of the first studies to compare the two widely used deep learning models and demonstrate the superior performance of the RNN model for clinical NER. PMID:29854252

Deep-cascade: Cascading 3D Deep Neural Networks for Fast Anomaly Detection and Localization in Crowded Scenes.

PubMed

Sabokrou, Mohammad; Fayyaz, Mohsen; Fathy, Mahmood; Klette, Reinhard

2017-02-17

This paper proposes a fast and reliable method for anomaly detection and localization in video data showing crowded scenes. Time-efficient anomaly localization is an ongoing challenge and subject of this paper. We propose a cubicpatch- based method, characterised by a cascade of classifiers, which makes use of an advanced feature-learning approach. Our cascade of classifiers has two main stages. First, a light but deep 3D auto-encoder is used for early identification of "many" normal cubic patches. This deep network operates on small cubic patches as being the first stage, before carefully resizing remaining candidates of interest, and evaluating those at the second stage using a more complex and deeper 3D convolutional neural network (CNN). We divide the deep autoencoder and the CNN into multiple sub-stages which operate as cascaded classifiers. Shallow layers of the cascaded deep networks (designed as Gaussian classifiers, acting as weak single-class classifiers) detect "simple" normal patches such as background patches, and more complex normal patches are detected at deeper layers. It is shown that the proposed novel technique (a cascade of two cascaded classifiers) performs comparable to current top-performing detection and localization methods on standard benchmarks, but outperforms those in general with respect to required computation time.

Learning representations for the early detection of sepsis with deep neural networks.

PubMed

Kam, Hye Jin; Kim, Ha Young

2017-10-01

Sepsis is one of the leading causes of death in intensive care unit patients. Early detection of sepsis is vital because mortality increases as the sepsis stage worsens. This study aimed to develop detection models for the early stage of sepsis using deep learning methodologies, and to compare the feasibility and performance of the new deep learning methodology with those of the regression method with conventional temporal feature extraction. Study group selection adhered to the InSight model. The results of the deep learning-based models and the InSight model were compared. With deep feedforward networks, the area under the ROC curve (AUC) of the models were 0.887 and 0.915 for the InSight and the new feature sets, respectively. For the model with the combined feature set, the AUC was the same as that of the basic feature set (0.915). For the long short-term memory model, only the basic feature set was applied and the AUC improved to 0.929 compared with the existing 0.887 of the InSight model. The contributions of this paper can be summarized in three ways: (i) improved performance without feature extraction using domain knowledge, (ii) verification of feature extraction capability of deep neural networks through comparison with reference features, and (iii) improved performance with feedforward neural networks using long short-term memory, a neural network architecture that can learn sequential patterns. Copyright © 2017 Elsevier Ltd. All rights reserved.

Applications of deep convolutional neural networks to digitized natural history collections.

PubMed

Schuettpelz, Eric; Frandsen, Paul B; Dikow, Rebecca B; Brown, Abel; Orli, Sylvia; Peters, Melinda; Metallo, Adam; Funk, Vicki A; Dorr, Laurence J

2017-01-01

Natural history collections contain data that are critical for many scientific endeavors. Recent efforts in mass digitization are generating large datasets from these collections that can provide unprecedented insight. Here, we present examples of how deep convolutional neural networks can be applied in analyses of imaged herbarium specimens. We first demonstrate that a convolutional neural network can detect mercury-stained specimens across a collection with 90% accuracy. We then show that such a network can correctly distinguish two morphologically similar plant families 96% of the time. Discarding the most challenging specimen images increases accuracy to 94% and 99%, respectively. These results highlight the importance of mass digitization and deep learning approaches and reveal how they can together deliver powerful new investigative tools.

Background rejection in NEXT using deep neural networks

DOE PAGES

Renner, J.; Farbin, A.; Vidal, J. Muñoz; ...

2017-01-16

Here, we investigate the potential of using deep learning techniques to reject background events in searches for neutrinoless double beta decay with high pressure xenon time projection chambers capable of detailed track reconstruction. The differences in the topological signatures of background and signal events can be learned by deep neural networks via training over many thousands of events. These networks can then be used to classify further events as signal or background, providing an additional background rejection factor at an acceptable loss of efficiency. The networks trained in this study performed better than previous methods developed based on the usemore » of the same topological signatures by a factor of 1.2 to 1.6, and there is potential for further improvement.« less

Characterization of bacterial diversity associated with deep sea ferromanganese nodules from the South China Sea.

PubMed

Zhang, De-Chao; Liu, Yan-Xia; Li, Xin-Zheng

2015-09-01

Deep sea ferromanganese (FeMn) nodules contain metallic mineral resources and have great economic potential. In this study, a combination of culture-dependent and culture-independent (16S rRNA genes clone library and pyrosequencing) methods was used to investigate the bacterial diversity in FeMn nodules from Jiaolong Seamount, the South China Sea. Eleven bacterial strains including some moderate thermophiles were isolated. The majority of strains belonged to the phylum Proteobacteria; one isolate belonged to the phylum Firmicutes. A total of 259 near full-length bacterial 16S rRNA gene sequences in a clone library and 67,079 valid reads obtained using pyrosequencing indicated that members of the Gammaproteobacteria dominated, with the most abundant bacterial genera being Pseudomonas and Alteromonas. Sequence analysis indicated the presence of many organisms whose closest relatives are known manganese oxidizers, iron reducers, hydrogen-oxidizing bacteria and methylotrophs. This is the first reported investigation of bacterial diversity associated with deep sea FeMn nodules from the South China Sea.

Explaining and improving breast cancer information acquisition among African American women in the Deep South.

PubMed

Anderson-Lewis, Charkarra; Ross, Levi; Johnson, Jarrett; Hastrup, Janice L; Green, B Lee; Kohler, Connie L

2012-06-01

A major challenge facing contemporary cancer educators is how to optimize the dissemination of breast cancer prevention and control information to African American women in the Deep South who are believed to be cancer free. The purpose of this research was to provide insight into the breast cancer information-acquisition experiences of African American women in Alabama and Mississippi and to make recommendations on ways to better reach members of this high-risk, underserved population. Focus group methodology was used in a repeated, cross-sectional research design with 64 African American women, 35 years old or older who lived in one of four urban or rural counties in Alabama and Mississippi. Axial-coded themes emerged around sources of cancer information, patterns of information acquisition, characteristics of preferred sources, and characteristics of least-preferred sources. It is important to invest in lay health educators to optimize the dissemination of breast cancer information to African American women who are believed to be cancer free in the Deep South.

CAPILLARY NETWORK ANOMALIES IN BRANCH RETINAL VEIN OCCLUSION ON OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY.

PubMed

Rispoli, Marco; Savastano, Maria Cristina; Lumbroso, Bruno

2015-11-01

To analyze the foveal microvasculature features in eyes with branch retinal vein occlusion (BRVO) using optical coherence tomography angiography based on split spectrum amplitude decorrelation angiography technology. A total of 10 BRVO eyes (mean age 64.2 ± 8.02 range between 52 years and 76 years) were evaluated by optical coherence tomography angiography (XR-Avanti; Optovue). The macular angiography scan protocol covered a 3 mm × 3 mm area. The focus of angiography analysis were two retinal layers: superficial vascular network and deep vascular network. The following vascular morphological congestion parameters were assessed in the vein occlusion area in both the superficial and deep networks: foveal avascular zone enlargement, capillary non-perfusion occurrence, microvascular abnormalities appearance, and vascular congestion signs. Image analyses were performed by 2 masked observers and interobserver agreement of image analyses was 0.90 (κ = 0.225, P < 0.01). In both superficial and deep network of BRVO, a decrease in capillary density with foveal avascular zone enlargement, capillary non-perfusion occurrence, and microvascular abnormalities appearance was observed (P < 0.01). The deep network showed the main vascular congestion at the boundary between healthy and nonperfused retina. Optical coherence tomography angiography in BRVO allows to detect foveal avascular zone enlargement, capillary nonperfusion, microvascular abnormalities, and vascular congestion signs both in the superficial and deep capillary network in all eyes. Optical coherence tomography angiography technology is a potential clinical tool for BRVO diagnosis and follow-up, providing stratigraphic vascular details that have not been previously observed by standard fluorescein angiography. The normal retinal vascular nets and areas of nonperfusion and congestion can be identified at various retinal levels. Optical coherence tomography angiography provides noninvasive images of the retinal capillaries and vascular networks.

Deep water characteristics and circulation in the South China Sea

NASA Astrophysics Data System (ADS)

Wang, Aimei; Du, Yan; Peng, Shiqiu; Liu, Kexiu; Huang, Rui Xin

2018-04-01

This study investigates the deep circulation in the South China Sea (SCS) using oceanographic observations combined with results from a bottom layer reduced gravity model. The SCS water, 2000 m below the surface, is quite different from that in the adjacent Pacific Ocean, and it is characterized by its low dissolved oxygen (DO), high temperature and low salinity. The horizontal distribution of deep water properties indicates a basin-scale cyclonic circulation driven by the Luzon overflow. The results of the bottom layer reduced gravity model are consistent with the existence of the cyclonic circulation in the deep SCS. The circulation is stronger at the northern/western boundary. After overflowing the sill of the Luzon Strait, the deep water moves broadly southwestward, constrained by the 3500 m isobath. The broadening of the southward flow is induced by the downwelling velocity in the interior of the deep basin. The main deep circulation bifurcates into two branches after the Zhongsha Islands. The southward branch continues flowing along the 3500 m isobath, and the eastward branch forms the sub-basin scale cyclonic circulation around the seamounts in the central deep SCS. The returning flow along the east boundary is fairly weak. The numerical experiments of the bottom layer reduced gravity model reveal the important roles of topography, bottom friction, and the upwelling/downwelling pattern in controlling the spatial structure, particularly the strong, deep western boundary current.

Deep Space Network Antenna Monitoring Using Adaptive Time Series Methods and Hidden Markov Models

NASA Technical Reports Server (NTRS)

Smyth, Padhraic; Mellstrom, Jeff

1993-01-01

The Deep Space Network (DSN)(designed and operated by the Jet Propulsion Laboratory for the National Aeronautics and Space Administration (NASA) provides end-to-end telecommunication capabilities between earth and various interplanetary spacecraft throughout the solar system.

The deep space network, volume 19

NASA Technical Reports Server (NTRS)

1974-01-01

The progress is reported in the DSN for Nov. and Dec. 1973. Research is described for the following areas: functions and facilities, mission support for flight projects, tracking and ground-based navigation, spacecraft/ground communication, network control and operations technology, and deep space stations.

The factors affecting the development of national identity as South korean in north korean refugees living in South Korea.

PubMed

Yu, Shi-Eun; Eom, Jin-Sup; Jeon, Woo-Taek

2012-09-01

This study aims to observe the factors that influence the development of national identity of North Korean refugees who have resettled in South Korea. The study population was comprised of 500 North Korean refugees who immigrated to South Korea in 2007. The variables measured national identity as South Korean, a scale for discrimination perceived during daily life, a social for supporting social network, a for childhood trauma experience, traumatic experiences in North Korea, and traumatic experiences during the escape process. Factor analysis was conducted on the result from the scale for national identity as South Korean which produced 4 factors including national consciousness, positive emotions, positive values, and negative values. Multiple regression was done to identify how variables such as demographic data, discrimination, social network, and past trauma had influenced each of 4 factors. National identity was negatively related by traumatic experience during childhood, perceived discrimination, and positively influenced by social networks. Positive emotion was related negatively to education level in North Korea and perceived discrimination, but positively related to traumatic experiences in North Korea. Negative value was related positively age and perceived discrimination but negatively related to supporting social network. The results of this study suggests that promoting social networks, decreasing discrimination and healing past traumas were important factors for North Korean refugees in South Korea to facilitate a new national identity as a South Korean.

Problem-Based Learning to Foster Deep Learning in Preservice Geography Teacher Education

ERIC Educational Resources Information Center

Golightly, Aubrey; Raath, Schalk

2015-01-01

In South Africa, geography education students' approach to deep learning has received little attention. Therefore the purpose of this one-shot experimental case study was to evaluate the extent to which first-year geography education students used deep or surface learning in an embedded problem-based learning (PBL) format. The researchers measured…

Holography as deep learning

NASA Astrophysics Data System (ADS)

Gan, Wen-Cong; Shu, Fu-Wen

Quantum many-body problem with exponentially large degrees of freedom can be reduced to a tractable computational form by neural network method [G. Carleo and M. Troyer, Science 355 (2017) 602, arXiv:1606.02318.] The power of deep neural network (DNN) based on deep learning is clarified by mapping it to renormalization group (RG), which may shed lights on holographic principle by identifying a sequence of RG transformations to the AdS geometry. In this paper, we show that any network which reflects RG process has intrinsic hyperbolic geometry, and discuss the structure of entanglement encoded in the graph of DNN. We find the entanglement structure of DNN is of Ryu-Takayanagi form. Based on these facts, we argue that the emergence of holographic gravitational theory is related to deep learning process of the quantum-field theory.

Automated Melanoma Recognition in Dermoscopy Images via Very Deep Residual Networks.

PubMed

Yu, Lequan; Chen, Hao; Dou, Qi; Qin, Jing; Heng, Pheng-Ann

2017-04-01

Automated melanoma recognition in dermoscopy images is a very challenging task due to the low contrast of skin lesions, the huge intraclass variation of melanomas, the high degree of visual similarity between melanoma and non-melanoma lesions, and the existence of many artifacts in the image. In order to meet these challenges, we propose a novel method for melanoma recognition by leveraging very deep convolutional neural networks (CNNs). Compared with existing methods employing either low-level hand-crafted features or CNNs with shallower architectures, our substantially deeper networks (more than 50 layers) can acquire richer and more discriminative features for more accurate recognition. To take full advantage of very deep networks, we propose a set of schemes to ensure effective training and learning under limited training data. First, we apply the residual learning to cope with the degradation and overfitting problems when a network goes deeper. This technique can ensure that our networks benefit from the performance gains achieved by increasing network depth. Then, we construct a fully convolutional residual network (FCRN) for accurate skin lesion segmentation, and further enhance its capability by incorporating a multi-scale contextual information integration scheme. Finally, we seamlessly integrate the proposed FCRN (for segmentation) and other very deep residual networks (for classification) to form a two-stage framework. This framework enables the classification network to extract more representative and specific features based on segmented results instead of the whole dermoscopy images, further alleviating the insufficiency of training data. The proposed framework is extensively evaluated on ISBI 2016 Skin Lesion Analysis Towards Melanoma Detection Challenge dataset. Experimental results demonstrate the significant performance gains of the proposed framework, ranking the first in classification and the second in segmentation among 25 teams and 28 teams, respectively. This study corroborates that very deep CNNs with effective training mechanisms can be employed to solve complicated medical image analysis tasks, even with limited training data.

Distribution and Diversity of Microbial Eukaryotes in Bathypelagic Waters of the South China Sea.

PubMed

Xu, Dapeng; Jiao, Nianzhi; Ren, Rui; Warren, Alan

2017-05-01

Little is known about the biodiversity of microbial eukaryotes in the South China Sea, especially in waters at bathyal depths. Here, we employed SSU rDNA gene sequencing to reveal the diversity and community structure across depth and distance gradients in the South China Sea. Vertically, the highest alpha diversity was found at 75-m depth. The communities of microbial eukaryotes were clustered into shallow-, middle-, and deep-water groups according to the depth from which they were collected, indicating a depth-related diversity and distribution pattern. Rhizaria sequences dominated the microeukaryote community and occurred in all samples except those from less than 50-m deep, being most abundant near the sea floor where they contributed ca. 64-97% and 40-74% of the total sequences and OTUs recovered, respectively. A large portion of rhizarian OTUs has neither a nearest named neighbor nor a nearest neighbor in the GenBank database which indicated the presence of new phylotypes in the South China Sea. Given their overwhelming abundance and richness, further phylogenetic analysis of rhizarians were performed and three new genetic clusters were revealed containing sequences retrieved from the deep waters of the South China Sea. Our results shed light on the diversity and community structure of microbial eukaryotes in this not yet fully explored area. © 2016 The Author(s) Journal of Eukaryotic Microbiology © 2016 International Society of Protistologists.

An analysis of image storage systems for scalable training of deep neural networks

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

Lim, Seung-Hwan; Young, Steven R; Patton, Robert M

This study presents a principled empirical evaluation of image storage systems for training deep neural networks. We employ the Caffe deep learning framework to train neural network models for three different data sets, MNIST, CIFAR-10, and ImageNet. While training the models, we evaluate five different options to retrieve training image data: (1) PNG-formatted image files on local file system; (2) pushing pixel arrays from image files into a single HDF5 file on local file system; (3) in-memory arrays to hold the pixel arrays in Python and C++; (4) loading the training data into LevelDB, a log-structured merge tree based key-valuemore » storage; and (5) loading the training data into LMDB, a B+tree based key-value storage. The experimental results quantitatively highlight the disadvantage of using normal image files on local file systems to train deep neural networks and demonstrate reliable performance with key-value storage based storage systems. When training a model on the ImageNet dataset, the image file option was more than 17 times slower than the key-value storage option. Along with measurements on training time, this study provides in-depth analysis on the cause of performance advantages/disadvantages of each back-end to train deep neural networks. We envision the provided measurements and analysis will shed light on the optimal way to architect systems for training neural networks in a scalable manner.« less

Multi-level deep supervised networks for retinal vessel segmentation.

PubMed

Mo, Juan; Zhang, Lei

2017-12-01

Changes in the appearance of retinal blood vessels are an important indicator for various ophthalmologic and cardiovascular diseases, including diabetes, hypertension, arteriosclerosis, and choroidal neovascularization. Vessel segmentation from retinal images is very challenging because of low blood vessel contrast, intricate vessel topology, and the presence of pathologies such as microaneurysms and hemorrhages. To overcome these challenges, we propose a neural network-based method for vessel segmentation. A deep supervised fully convolutional network is developed by leveraging multi-level hierarchical features of the deep networks. To improve the discriminative capability of features in lower layers of the deep network and guide the gradient back propagation to overcome gradient vanishing, deep supervision with auxiliary classifiers is incorporated in some intermediate layers of the network. Moreover, the transferred knowledge learned from other domains is used to alleviate the issue of insufficient medical training data. The proposed approach does not rely on hand-crafted features and needs no problem-specific preprocessing or postprocessing, which reduces the impact of subjective factors. We evaluate the proposed method on three publicly available databases, the DRIVE, STARE, and CHASE_DB1 databases. Extensive experiments demonstrate that our approach achieves better or comparable performance to state-of-the-art methods with a much faster processing speed, making it suitable for real-world clinical applications. The results of cross-training experiments demonstrate its robustness with respect to the training set. The proposed approach segments retinal vessels accurately with a much faster processing speed and can be easily applied to other biomedical segmentation tasks.

Pattern learning with deep neural networks in EMG-based speech recognition.

PubMed

Wand, Michael; Schultz, Tanja

2014-01-01

We report on classification of phones and phonetic features from facial electromyographic (EMG) data, within the context of our EMG-based Silent Speech interface. In this paper we show that a Deep Neural Network can be used to perform this classification task, yielding a significant improvement over conventional Gaussian Mixture models. Our central contribution is the visualization of patterns which are learned by the neural network. With increasing network depth, these patterns represent more and more intricate electromyographic activity.

Neural network based satellite tracking for deep space applications

NASA Technical Reports Server (NTRS)

Amoozegar, F.; Ruggier, C.

2003-01-01

The objective of this paper is to provide a survey of neural network trends as applied to the tracking of spacecrafts in deep space at Ka-band under various weather conditions and examine the trade-off between tracing accuracy and communication link performance.

High-power transmitter automation. [deep space network

NASA Technical Reports Server (NTRS)

Gosline, R.

1980-01-01

The current status of the transmitter automation development applicable to all transmitters in the deep space network is described. Interface and software designs are described that improve reliability and reduce the time required for subsystem turn-on and klystron saturation to less than 10 minutes.

Convolutional networks for fast, energy-efficient neuromorphic computing

PubMed Central

Esser, Steven K.; Merolla, Paul A.; Arthur, John V.; Cassidy, Andrew S.; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J.; McKinstry, Jeffrey L.; Melano, Timothy; Barch, Davis R.; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D.; Modha, Dharmendra S.

2016-01-01

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware’s underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer. PMID:27651489

Convolutional networks for fast, energy-efficient neuromorphic computing.

PubMed

Esser, Steven K; Merolla, Paul A; Arthur, John V; Cassidy, Andrew S; Appuswamy, Rathinakumar; Andreopoulos, Alexander; Berg, David J; McKinstry, Jeffrey L; Melano, Timothy; Barch, Davis R; di Nolfo, Carmelo; Datta, Pallab; Amir, Arnon; Taba, Brian; Flickner, Myron D; Modha, Dharmendra S

2016-10-11

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.

Shakeout: A New Approach to Regularized Deep Neural Network Training.

PubMed

Kang, Guoliang; Li, Jun; Tao, Dacheng

2018-05-01

Recent years have witnessed the success of deep neural networks in dealing with a plenty of practical problems. Dropout has played an essential role in many successful deep neural networks, by inducing regularization in the model training. In this paper, we present a new regularized training approach: Shakeout. Instead of randomly discarding units as Dropout does at the training stage, Shakeout randomly chooses to enhance or reverse each unit's contribution to the next layer. This minor modification of Dropout has the statistical trait: the regularizer induced by Shakeout adaptively combines , and regularization terms. Our classification experiments with representative deep architectures on image datasets MNIST, CIFAR-10 and ImageNet show that Shakeout deals with over-fitting effectively and outperforms Dropout. We empirically demonstrate that Shakeout leads to sparser weights under both unsupervised and supervised settings. Shakeout also leads to the grouping effect of the input units in a layer. Considering the weights in reflecting the importance of connections, Shakeout is superior to Dropout, which is valuable for the deep model compression. Moreover, we demonstrate that Shakeout can effectively reduce the instability of the training process of the deep architecture.

Deep circulation changes in the South Atlantic since the Last Glacial Maximum from Nd isotope and multi-proxy records

NASA Astrophysics Data System (ADS)

Wei, R.; Abouchami, W.; Zahn, R.; Masque, P.

2016-01-01

We report down-core sedimentary Nd isotope (εNd) records from two South Atlantic sediment cores, MD02-2594 and GeoB3603-2, located on the western South African continental margin. The core sites are positioned downstream of the present-day flow path of North Atlantic Deep Water (NADW) and close to the Southern Ocean, which makes them suitable for reconstructing past variability in NADW circulation over the last glacial cycle. The Fe-Mn leachates εNd records show a coherent decreasing trend from glacial radiogenic values towards less radiogenic values during the Holocene. This trend is confirmed by εNd in fish debris and mixed planktonic foraminifera, albeit with an offset during the Holocene to lower values relative to the leachates, matching the present-day composition of NADW in the Cape Basin. We interpret the εNd changes as reflecting the glacial shoaling of Southern Ocean waters to shallower depths combined with the admixing of southward flowing Northern Component Water (NCW). A compilation of Atlantic εNd records reveals increasing radiogenic isotope signatures towards the south and with increasing depth. This signal is most prominent during the Last Glacial Maximum (LGM) and of similar amplitude across the Atlantic basin, suggesting continuous deep water production in the North Atlantic and export to the South Atlantic and the Southern Ocean. The amplitude of the εNd change from the LGM to Holocene is largest in the southernmost cores, implying a greater sensitivity to the deglacial strengthening of NADW at these sites. This signal impacted most prominently the South Atlantic deep and bottom water layers that were particularly deprived of NCW during the LGM. The εNd variations correlate with changes in 231Pa/230Th ratios and benthic δ13C across the deglacial transition. Together with the contrasting 231Pa/230Th: εNd pattern of the North and South Atlantic, this indicates a progressive reorganization of the AMOC to full strength during the Holocene.

SYNAPTIC DEPRESSION IN DEEP NEURAL NETWORKS FOR SPEECH PROCESSING.

PubMed

Zhang, Wenhao; Li, Hanyu; Yang, Minda; Mesgarani, Nima

2016-03-01

A characteristic property of biological neurons is their ability to dynamically change the synaptic efficacy in response to variable input conditions. This mechanism, known as synaptic depression, significantly contributes to the formation of normalized representation of speech features. Synaptic depression also contributes to the robust performance of biological systems. In this paper, we describe how synaptic depression can be modeled and incorporated into deep neural network architectures to improve their generalization ability. We observed that when synaptic depression is added to the hidden layers of a neural network, it reduces the effect of changing background activity in the node activations. In addition, we show that when synaptic depression is included in a deep neural network trained for phoneme classification, the performance of the network improves under noisy conditions not included in the training phase. Our results suggest that more complete neuron models may further reduce the gap between the biological performance and artificial computing, resulting in networks that better generalize to novel signal conditions.

Deep learning methods for protein torsion angle prediction.

PubMed

Li, Haiou; Hou, Jie; Adhikari, Badri; Lyu, Qiang; Cheng, Jianlin

2017-09-18

Deep learning is one of the most powerful machine learning methods that has achieved the state-of-the-art performance in many domains. Since deep learning was introduced to the field of bioinformatics in 2012, it has achieved success in a number of areas such as protein residue-residue contact prediction, secondary structure prediction, and fold recognition. In this work, we developed deep learning methods to improve the prediction of torsion (dihedral) angles of proteins. We design four different deep learning architectures to predict protein torsion angles. The architectures including deep neural network (DNN) and deep restricted Boltzmann machine (DRBN), deep recurrent neural network (DRNN) and deep recurrent restricted Boltzmann machine (DReRBM) since the protein torsion angle prediction is a sequence related problem. In addition to existing protein features, two new features (predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments) are used as input to each of the four deep learning architectures to predict phi and psi angles of protein backbone. The mean absolute error (MAE) of phi and psi angles predicted by DRNN, DReRBM, DRBM and DNN is about 20-21° and 29-30° on an independent dataset. The MAE of phi angle is comparable to the existing methods, but the MAE of psi angle is 29°, 2° lower than the existing methods. On the latest CASP12 targets, our methods also achieved the performance better than or comparable to a state-of-the art method. Our experiment demonstrates that deep learning is a valuable method for predicting protein torsion angles. The deep recurrent network architecture performs slightly better than deep feed-forward architecture, and the predicted residue contact number and the error distribution of torsion angles extracted from sequence fragments are useful features for improving prediction accuracy.

Trace element evidence for abrupt changes in deep South Atlantic Ocean nutrient and carbonate chemistry across the Mid-Pleistocene Transition

NASA Astrophysics Data System (ADS)

Farmer, J. R.; Hoenisch, B.; Haynes, L.; Kroon, D.; Bell, D. B.; Jung, S.; Seguí, M. J.; Raymo, M. E.; Goldstein, S. L.; Pena, L. D.

2016-12-01

Pleistocene glaciations underwent a profound transition from lower amplitude 40 kyr cycles to high amplitude 100 kyr cycles between 1.2 and 0.8 Ma, an interval termed the Mid-Pleistocene Transition (MPT). While the underlying causes of the MPT are uncertain, previous studies show quasi-contemporaneous reductions in North Atlantic Deep Water (NADW) export1 and glacial atmospheric pCO22 around 0.9 Ma. Although this suggests a possible role for enhanced deep-ocean carbon storage in amplifying climate change across the MPT, few direct records of deep ocean carbonate chemistry exist for this interval to test this hypothesis. Here we present South Atlantic benthic foraminiferal B/Ca and Cd/Ca records from International Ocean Discovery Program Sites 1088, 1264 and 1267 (2.1 to 4.3 km water depth) as part of a larger study of Atlantic-wide changes in deep ocean chemistry and circulation spanning the MPT. Results show an abrupt 15-20% decrease in benthic B/Ca and 40-50% increase in Cd/Ca at 4.3 km depth (Site 1267) between 1.0 and 0.9 Ma. Site 1088, which at 2.1 km depth is sensitive to input of southern-sourced Upper Circumpolar Deep Water, shows a prolonged 25% decrease in B/Ca and 50% increase in Cd/Ca from 1.0 to 0.6 Ma. In contrast, at Site 1264 ( 2.5 km depth within the core of modern NADW) B/Ca and Cd/Ca changes across the MPT are more modest (-5% and +10%, respectively). These observations reflect on the accumulation of regenerated carbon and nutrients in the deep South Atlantic, and varying contributions of northern- and southern-sourced watermasses to each core site. Implications for deep-ocean carbon storage and forcing of the MPT will be discussed. 1Pena, L. and Goldstein, S. (2014), Science 345, 318 2Hönisch, B. et al. (2009), Science 324, 1551

Deep Learning for Computer Vision: A Brief Review

PubMed Central

Doulamis, Nikolaos; Doulamis, Anastasios; Protopapadakis, Eftychios

2018-01-01

Over the last years deep learning methods have been shown to outperform previous state-of-the-art machine learning techniques in several fields, with computer vision being one of the most prominent cases. This review paper provides a brief overview of some of the most significant deep learning schemes used in computer vision problems, that is, Convolutional Neural Networks, Deep Boltzmann Machines and Deep Belief Networks, and Stacked Denoising Autoencoders. A brief account of their history, structure, advantages, and limitations is given, followed by a description of their applications in various computer vision tasks, such as object detection, face recognition, action and activity recognition, and human pose estimation. Finally, a brief overview is given of future directions in designing deep learning schemes for computer vision problems and the challenges involved therein. PMID:29487619

DCMDN: Deep Convolutional Mixture Density Network

NASA Astrophysics Data System (ADS)

D'Isanto, Antonio; Polsterer, Kai Lars

2017-09-01

Deep Convolutional Mixture Density Network (DCMDN) estimates probabilistic photometric redshift directly from multi-band imaging data by combining a version of a deep convolutional network with a mixture density network. The estimates are expressed as Gaussian mixture models representing the probability density functions (PDFs) in the redshift space. In addition to the traditional scores, the continuous ranked probability score (CRPS) and the probability integral transform (PIT) are applied as performance criteria. DCMDN is able to predict redshift PDFs independently from the type of source, e.g. galaxies, quasars or stars and renders pre-classification of objects and feature extraction unnecessary; the method is extremely general and allows the solving of any kind of probabilistic regression problems based on imaging data, such as estimating metallicity or star formation rate in galaxies.

Applications of deep convolutional neural networks to digitized natural history collections

PubMed Central

Frandsen, Paul B.; Dikow, Rebecca B.; Brown, Abel; Orli, Sylvia; Peters, Melinda; Metallo, Adam; Funk, Vicki A.; Dorr, Laurence J.

2017-01-01

Abstract Natural history collections contain data that are critical for many scientific endeavors. Recent efforts in mass digitization are generating large datasets from these collections that can provide unprecedented insight. Here, we present examples of how deep convolutional neural networks can be applied in analyses of imaged herbarium specimens. We first demonstrate that a convolutional neural network can detect mercury-stained specimens across a collection with 90% accuracy. We then show that such a network can correctly distinguish two morphologically similar plant families 96% of the time. Discarding the most challenging specimen images increases accuracy to 94% and 99%, respectively. These results highlight the importance of mass digitization and deep learning approaches and reveal how they can together deliver powerful new investigative tools. PMID:29200929

Maximum entropy methods for extracting the learned features of deep neural networks.

PubMed

Finnegan, Alex; Song, Jun S

2017-10-01

New architectures of multilayer artificial neural networks and new methods for training them are rapidly revolutionizing the application of machine learning in diverse fields, including business, social science, physical sciences, and biology. Interpreting deep neural networks, however, currently remains elusive, and a critical challenge lies in understanding which meaningful features a network is actually learning. We present a general method for interpreting deep neural networks and extracting network-learned features from input data. We describe our algorithm in the context of biological sequence analysis. Our approach, based on ideas from statistical physics, samples from the maximum entropy distribution over possible sequences, anchored at an input sequence and subject to constraints implied by the empirical function learned by a network. Using our framework, we demonstrate that local transcription factor binding motifs can be identified from a network trained on ChIP-seq data and that nucleosome positioning signals are indeed learned by a network trained on chemical cleavage nucleosome maps. Imposing a further constraint on the maximum entropy distribution also allows us to probe whether a network is learning global sequence features, such as the high GC content in nucleosome-rich regions. This work thus provides valuable mathematical tools for interpreting and extracting learned features from feed-forward neural networks.

Depleted deep South China Sea δ13C paleoceanographic events in response to tectonic evolution in Taiwan-Luzon Strait since Middle Miocene

NASA Astrophysics Data System (ADS)

Chen, Wen-Huang; Huang, Chi-Yue; Lin, Yen-Jun; Zhao, Quanhong; Yan, Yi; Chen, Duofu; Zhang, Xinchang; Lan, Qing; Yu, Mengming

2015-12-01

The most distinctive feature of the deep South China Sea (SCS) paleoceanography is the occurrence of long-term depleted deep-sea benthic foraminiferal δ13C values. They are lower than the global and the Pacific composite records in the last 16 Ma, especially at 13.2, 10.5, 6.5, 3.0 and 1.2-0.4 Ma. This distinct deep SCS paleoceanograhic history coincides with the subduction-collision history in the Taiwan region where waters of the West Pacific (WP) and the SCS exchange. The depleted deep-sea benthic foraminiferal δ13C events indicate that the SCS deep basin became progressively a stagnant environment in the last 16 Ma due to either closure of the connection with the WP bottom water or temporary reduction of the WP deep water flowing into the deep SCS. Both the Taiwan accretionary prism and the Luzon arc became the main tectono-morphological barriers for the WP bottom water flowing into the SCS deep basin when eastward subduction of the SCS oceanic lithosphere beneath the Philippine Sea Plate started from the Middle Miocene (18-16 Ma). This began a long-term trend of depleted SCS deep-sea benthic δ13C values in the last 16 Ma. The oblique arc-continent collision since ~6.5 Ma uplifted the Taiwan accretionary prism rapidly above sea level and further isolated the SCS from the open Pacific. The collision simultaneously causes backthrusting deformations in the North Luzon Trough forearc basin and sequentially closes interarc water gates between volcanic islands from north to south. The Loho and the Taitung interarc water gates in the advanced collision zone were closed at ~3.0 Ma and ~1.2 Ma, coinciding with the very low SCS deep-sea benthic δ13C events at 3.0 and 1.2-0.4 Ma, respectively. The Taitung Canyon between the Lutao and Lanyu volcanic islands in the incipient collision zone is semi-closed presently. These closure events also lead to the result that the WP deep water intrudes westward into the SCS principally through the Bashi Channel between the Lanyu and Batan volcanic islands in the subduction zone.

A federated information management system for the Deep Space Network. M.S. Thesis - Univ. of Southern California

NASA Technical Reports Server (NTRS)

Dobinson, E.

1982-01-01

General requirements for an information management system for the deep space network (DSN) are examined. A concise review of available database management system technology is presented. It is recommended that a federation of logically decentralized databases be implemented for the Network Information Management System of the DSN. Overall characteristics of the federation are specified, as well as reasons for adopting this approach.

Effect of Wave Boundary Layer on Sea-to-Air Dimethylsulfide Transfer Velocity During Typhoon Passage

DTIC Science & Technology

2006-09-01

straits are shallow except Luzon Strait whose maximum depth is 1800 m. The elliptical shaped central deep basin is 1900 km along its major axis...the east, Borneo to the southeast, and Indonesia to the south, a total area of 3.5·106 km2. Its southern border is 3° S between South Sumatra and...averaging 100 m deep on the continental Sunda shelf and over 5000 m in the Philippine basin ; and its unusual monsoon weather patterns of reversing

ELLICOTT ROCK WILDERNESS AND ADDITIONS, SOUTH CAROLINA, NORTH CAROLINA, AND GEORGIA.

USGS Publications Warehouse

Bell, Henry; Gazdik, Gertrude C.

1984-01-01

A mineral survey was made of the Ellicott Rock Wilderness and additions located in the common corner of South Carolina, North Carolina, and Georgia. Surveys along the rivers, streams, and ridges indicated that there is little promise for the occurrence of metallic mineral or energy resources. Deeply buried sedimentary rocks have an unknown potential for hydrocarbons, probably gas. Until some deep drilling is done to test these deep sedimentary rocks no reasonable estimate of gas potential can be made, but it cannot be totally discounted.

ChemNet: A Transferable and Generalizable Deep Neural Network for Small-Molecule Property Prediction

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

Goh, Garrett B.; Siegel, Charles M.; Vishnu, Abhinav

With access to large datasets, deep neural networks through representation learning have been able to identify patterns from raw data, achieving human-level accuracy in image and speech recognition tasks. However, in chemistry, availability of large standardized and labelled datasets is scarce, and with a multitude of chemical properties of interest, chemical data is inherently small and fragmented. In this work, we explore transfer learning techniques in conjunction with the existing Chemception CNN model, to create a transferable and generalizable deep neural network for small-molecule property prediction. Our latest model, ChemNet learns in a semi-supervised manner from inexpensive labels computed frommore » the ChEMBL database. When fine-tuned to the Tox21, HIV and FreeSolv dataset, which are 3 separate chemical tasks that ChemNet was not originally trained on, we demonstrate that ChemNet exceeds the performance of existing Chemception models, contemporary MLP models that trains on molecular fingerprints, and it matches the performance of the ConvGraph algorithm, the current state-of-the-art. Furthermore, as ChemNet has been pre-trained on a large diverse chemical database, it can be used as a universal “plug-and-play” deep neural network, which accelerates the deployment of deep neural networks for the prediction of novel small-molecule chemical properties.« less

Modeling language and cognition with deep unsupervised learning: a tutorial overview

PubMed Central

Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P.

2013-01-01

Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition. PMID:23970869

Deep convolutional neural network based antenna selection in multiple-input multiple-output system

NASA Astrophysics Data System (ADS)

Cai, Jiaxin; Li, Yan; Hu, Ying

2018-03-01

Antenna selection of wireless communication system has attracted increasing attention due to the challenge of keeping a balance between communication performance and computational complexity in large-scale Multiple-Input MultipleOutput antenna systems. Recently, deep learning based methods have achieved promising performance for large-scale data processing and analysis in many application fields. This paper is the first attempt to introduce the deep learning technique into the field of Multiple-Input Multiple-Output antenna selection in wireless communications. First, the label of attenuation coefficients channel matrix is generated by minimizing the key performance indicator of training antenna systems. Then, a deep convolutional neural network that explicitly exploits the massive latent cues of attenuation coefficients is learned on the training antenna systems. Finally, we use the adopted deep convolutional neural network to classify the channel matrix labels of test antennas and select the optimal antenna subset. Simulation experimental results demonstrate that our method can achieve better performance than the state-of-the-art baselines for data-driven based wireless antenna selection.

Modeling language and cognition with deep unsupervised learning: a tutorial overview.

PubMed

Zorzi, Marco; Testolin, Alberto; Stoianov, Ivilin P

2013-01-01

Deep unsupervised learning in stochastic recurrent neural networks with many layers of hidden units is a recent breakthrough in neural computation research. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. In this article we discuss the theoretical foundations of this approach and we review key issues related to training, testing and analysis of deep networks for modeling language and cognitive processing. The classic letter and word perception problem of McClelland and Rumelhart (1981) is used as a tutorial example to illustrate how structured and abstract representations may emerge from deep generative learning. We argue that the focus on deep architectures and generative (rather than discriminative) learning represents a crucial step forward for the connectionist modeling enterprise, because it offers a more plausible model of cortical learning as well as a way to bridge the gap between emergentist connectionist models and structured Bayesian models of cognition.

Action-Driven Visual Object Tracking With Deep Reinforcement Learning.

PubMed

Yun, Sangdoo; Choi, Jongwon; Yoo, Youngjoon; Yun, Kimin; Choi, Jin Young

2018-06-01

In this paper, we propose an efficient visual tracker, which directly captures a bounding box containing the target object in a video by means of sequential actions learned using deep neural networks. The proposed deep neural network to control tracking actions is pretrained using various training video sequences and fine-tuned during actual tracking for online adaptation to a change of target and background. The pretraining is done by utilizing deep reinforcement learning (RL) as well as supervised learning. The use of RL enables even partially labeled data to be successfully utilized for semisupervised learning. Through the evaluation of the object tracking benchmark data set, the proposed tracker is validated to achieve a competitive performance at three times the speed of existing deep network-based trackers. The fast version of the proposed method, which operates in real time on graphics processing unit, outperforms the state-of-the-art real-time trackers with an accuracy improvement of more than 8%.

Visit to the Deep Underground Science and Engineering Laboratory

ScienceCinema

None

2017-12-09

U.S. Department of Energy scientists and administrators join members of the National Science Foundation and South Dakotas Sanford Underground Laboratory for the deepest journey yet to the proposed site of the Deep Underground Science and Engineering Laboratory (DUSEL).

Visit to the Deep Underground Science and Engineering Laboratory

ScienceCinema

None

2018-05-16

U.S. Department of Energy scientists and administrators join members of the National Science Foundation and South Dakotas Sanford Underground Laboratory for the deepest journey yet to the proposed site of the Deep Underground Science and Engineering Laboratory (DUSEL).

Impact of deep learning on the normalization of reconstruction kernel effects in imaging biomarker quantification: a pilot study in CT emphysema

NASA Astrophysics Data System (ADS)

Jin, Hyeongmin; Heo, Changyong; Kim, Jong Hyo

2018-02-01

Differing reconstruction kernels are known to strongly affect the variability of imaging biomarkers and thus remain as a barrier in translating the computer aided quantification techniques into clinical practice. This study presents a deep learning application to CT kernel conversion which converts a CT image of sharp kernel to that of standard kernel and evaluates its impact on variability reduction of a pulmonary imaging biomarker, the emphysema index (EI). Forty cases of low-dose chest CT exams obtained with 120kVp, 40mAs, 1mm thickness, of 2 reconstruction kernels (B30f, B50f) were selected from the low dose lung cancer screening database of our institution. A Fully convolutional network was implemented with Keras deep learning library. The model consisted of symmetric layers to capture the context and fine structure characteristics of CT images from the standard and sharp reconstruction kernels. Pairs of the full-resolution CT data set were fed to input and output nodes to train the convolutional network to learn the appropriate filter kernels for converting the CT images of sharp kernel to standard kernel with a criterion of measuring the mean squared error between the input and target images. EIs (RA950 and Perc15) were measured with a software package (ImagePrism Pulmo, Seoul, South Korea) and compared for the data sets of B50f, B30f, and the converted B50f. The effect of kernel conversion was evaluated with the mean and standard deviation of pair-wise differences in EI. The population mean of RA950 was 27.65 +/- 7.28% for B50f data set, 10.82 +/- 6.71% for the B30f data set, and 8.87 +/- 6.20% for the converted B50f data set. The mean of pair-wise absolute differences in RA950 between B30f and B50f is reduced from 16.83% to 1.95% using kernel conversion. Our study demonstrates the feasibility of applying the deep learning technique for CT kernel conversion and reducing the kernel-induced variability of EI quantification. The deep learning model has a potential to improve the reliability of imaging biomarker, especially in evaluating the longitudinal changes of EI even when the patient CT scans were performed with different kernels.

The Factors Affecting the Development of National Identity as South Korean in North Korean Refugees Living in South Korea

PubMed Central

Yu, Shi-Eun; Eom, Jin-Sup

2012-01-01

Objective This study aims to observe the factors that influence the development of national identity of North Korean refugees who have resettled in South Korea. Methods The study population was comprised of 500 North Korean refugees who immigrated to South Korea in 2007. The variables measured national identity as South Korean, a scale for discrimination perceived during daily life, a social for supporting social network, a for childhood trauma experience, traumatic experiences in North Korea, and traumatic experiences during the escape process. Factor analysis was conducted on the result from the scale for national identity as South Korean which produced 4 factors including national consciousness, positive emotions, positive values, and negative values. Multiple regression was done to identify how variables such as demographic data, discrimination, social network, and past trauma had influenced each of 4 factors. Results National identity was negatively related by traumatic experience during childhood, perceived discrimination, and positively influenced by social networks. Positive emotion was related negatively to education level in North Korea and perceived discrimination, but positively related to traumatic experiences in North Korea. Negative value was related positively age and perceived discrimination but negatively related to supporting social network. Conclusion The results of this study suggests that promoting social networks, decreasing discrimination and healing past traumas were important factors for North Korean refugees in South Korea to facilitate a new national identity as a South Korean. PMID:22993518

Propagation Effects of Importance to the NASA/JPL Deep Space Network (DSN)

NASA Technical Reports Server (NTRS)

Slobin, Steve

1999-01-01

This paper presents Propagation Effects of Importance To The NASA/JPL Deep Space Network (DSN). The topics include: 1) DSN Antennas; 2) Deep Space Telecom Link Basics; 3) DSN Propagation Region of Interest; 4) Ka-Band Weather Effects Models and Examples; 5) Existing Goldstone Ka-Band Atmosphere Attenuation Model; 6) Existing Goldstone Atmosphere Noise Temperature Model; and 7) Ka-Band delta (G/T) Relative to Vacuum Condition. This paper summarizes the topics above.

A novel deep learning approach for classification of EEG motor imagery signals.

PubMed

Tabar, Yousef Rezaei; Halici, Ugur

2017-02-01

Signal classification is an important issue in brain computer interface (BCI) systems. Deep learning approaches have been used successfully in many recent studies to learn features and classify different types of data. However, the number of studies that employ these approaches on BCI applications is very limited. In this study we aim to use deep learning methods to improve classification performance of EEG motor imagery signals. In this study we investigate convolutional neural networks (CNN) and stacked autoencoders (SAE) to classify EEG Motor Imagery signals. A new form of input is introduced to combine time, frequency and location information extracted from EEG signal and it is used in CNN having one 1D convolutional and one max-pooling layers. We also proposed a new deep network by combining CNN and SAE. In this network, the features that are extracted in CNN are classified through the deep network SAE. The classification performance obtained by the proposed method on BCI competition IV dataset 2b in terms of kappa value is 0.547. Our approach yields 9% improvement over the winner algorithm of the competition. Our results show that deep learning methods provide better classification performance compared to other state of art approaches. These methods can be applied successfully to BCI systems where the amount of data is large due to daily recording.

A Deep Neural Network Model for Rainfall Estimation UsingPolarimetric WSR-88DP Radar Observations

NASA Astrophysics Data System (ADS)

Tan, H.; Chandra, C. V.; Chen, H.

2016-12-01

Rainfall estimation based on radar measurements has been an important topic for a few decades. Generally, radar rainfall estimation is conducted through parametric algorisms such as reflectivity-rainfall relation (i.e., Z-R relation). On the other hand, neural networks are developed for ground rainfall estimation based on radar measurements. This nonparametric method, which takes into account of both radar observations and rainfall measurements from ground rain gauges, has been demonstrated successfully for rainfall rate estimation. However, the neural network-based rainfall estimation is limited in practice due to the model complexity and structure, data quality, as well as different rainfall microphysics. Recently, the deep learning approach has been introduced in pattern recognition and machine learning areas. Compared to traditional neural networks, the deep learning based methodologies have larger number of hidden layers and more complex structure for data representation. Through a hierarchical learning process, the high level structured information and knowledge can be extracted automatically from low level features of the data. In this paper, we introduce a novel deep neural network model for rainfall estimation based on ground polarimetric radar measurements .The model is designed to capture the complex abstractions of radar measurements at different levels using multiple layers feature identification and extraction. The abstractions at different levels can be used independently or fused with other data resource such as satellite-based rainfall products and/or topographic data to represent the rain characteristics at certain location. In particular, the WSR-88DP radar and rain gauge data collected in Dallas - Fort Worth Metroplex and Florida are used extensively to train the model, and for demonstration purposes. Quantitative evaluation of the deep neural network based rainfall products will also be presented, which is based on an independent rain gauge network.

The 1968 Edcouch-Elsa High School Walkout: Chicano Student Activism in a South Texas Community

ERIC Educational Resources Information Center

Barrera, James B.

2004-01-01

A nonviolent school boycott by 192 Chicanola students in 1968 at Edcouch-Elsa high school in the Rio Grande Valley region of Deep South Texas is examined. This walkout was the first major Chicano student protest in South Texas, and was a product of the 1960s Chicano movement.

Microwave analog fiber-optic link for use in the deep space network

NASA Technical Reports Server (NTRS)

Logan, R. T., Jr.; Lutes, G. F.; Maleki, L.

1990-01-01

A novel fiber-optic system with dynamic range of up to 150 dB-Hz for transmission of microwave analog signals is described. The design, analysis, and laboratory evaluations of this system are reported, and potential applications in the NASA/JPL Deep Space Network are discussed.

Evolutionary Scheduler for the Deep Space Network

NASA Technical Reports Server (NTRS)

Guillaume, Alexandre; Lee, Seungwon; Wang, Yeou-Fang; Zheng, Hua; Chau, Savio; Tung, Yu-Wen; Terrile, Richard J.; Hovden, Robert

2010-01-01

A computer program assists human schedulers in satisfying, to the maximum extent possible, competing demands from multiple spacecraft missions for utilization of the transmitting/receiving Earth stations of NASA s Deep Space Network. The program embodies a concept of optimal scheduling to attain multiple objectives in the presence of multiple constraints.

Neptune Hot South Pole

NASA Image and Video Library

2007-09-18

These thermal images show a hot south pole on the planet Neptune. These warmer temperatures provide an avenue for methane to escape out of the deep atmosphere. The images were obtained with the Very Large Telescope in Chile Sept. 1 and 2, 2006.

14. Detail view of columns, capitals and beams at south ...

Library of Congress Historic Buildings Survey, Historic Engineering Record, Historic Landscapes Survey

14. Detail view of columns, capitals and beams at south end of north section of mill. Note the transition from deep pocket to shallow pocket column capitals. - Lowe Mill, Eighth Avenue, Southwest, Huntsville, Madison County, AL

Deep Visual Attention Prediction

NASA Astrophysics Data System (ADS)

Wang, Wenguan; Shen, Jianbing

2018-05-01

In this work, we aim to predict human eye fixation with view-free scenes based on an end-to-end deep learning architecture. Although Convolutional Neural Networks (CNNs) have made substantial improvement on human attention prediction, it is still needed to improve CNN based attention models by efficiently leveraging multi-scale features. Our visual attention network is proposed to capture hierarchical saliency information from deep, coarse layers with global saliency information to shallow, fine layers with local saliency response. Our model is based on a skip-layer network structure, which predicts human attention from multiple convolutional layers with various reception fields. Final saliency prediction is achieved via the cooperation of those global and local predictions. Our model is learned in a deep supervision manner, where supervision is directly fed into multi-level layers, instead of previous approaches of providing supervision only at the output layer and propagating this supervision back to earlier layers. Our model thus incorporates multi-level saliency predictions within a single network, which significantly decreases the redundancy of previous approaches of learning multiple network streams with different input scales. Extensive experimental analysis on various challenging benchmark datasets demonstrate our method yields state-of-the-art performance with competitive inference time.

On the complexity of neural network classifiers: a comparison between shallow and deep architectures.

PubMed

Bianchini, Monica; Scarselli, Franco

2014-08-01

Recently, researchers in the artificial neural network field have focused their attention on connectionist models composed by several hidden layers. In fact, experimental results and heuristic considerations suggest that deep architectures are more suitable than shallow ones for modern applications, facing very complex problems, e.g., vision and human language understanding. However, the actual theoretical results supporting such a claim are still few and incomplete. In this paper, we propose a new approach to study how the depth of feedforward neural networks impacts on their ability in implementing high complexity functions. First, a new measure based on topological concepts is introduced, aimed at evaluating the complexity of the function implemented by a neural network, used for classification purposes. Then, deep and shallow neural architectures with common sigmoidal activation functions are compared, by deriving upper and lower bounds on their complexity, and studying how the complexity depends on the number of hidden units and the used activation function. The obtained results seem to support the idea that deep networks actually implements functions of higher complexity, so that they are able, with the same number of resources, to address more difficult problems.

Cytopathological image analysis using deep-learning networks in microfluidic microscopy.

PubMed

Gopakumar, G; Hari Babu, K; Mishra, Deepak; Gorthi, Sai Siva; Sai Subrahmanyam, Gorthi R K

2017-01-01

Cytopathologic testing is one of the most critical steps in the diagnosis of diseases, including cancer. However, the task is laborious and demands skill. Associated high cost and low throughput drew considerable interest in automating the testing process. Several neural network architectures were designed to provide human expertise to machines. In this paper, we explore and propose the feasibility of using deep-learning networks for cytopathologic analysis by performing the classification of three important unlabeled, unstained leukemia cell lines (K562, MOLT, and HL60). The cell images used in the classification are captured using a low-cost, high-throughput cell imaging technique: microfluidics-based imaging flow cytometry. We demonstrate that without any conventional fine segmentation followed by explicit feature extraction, the proposed deep-learning algorithms effectively classify the coarsely localized cell lines. We show that the designed deep belief network as well as the deeply pretrained convolutional neural network outperform the conventionally used decision systems and are important in the medical domain, where the availability of labeled data is limited for training. We hope that our work enables the development of a clinically significant high-throughput microfluidic microscopy-based tool for disease screening/triaging, especially in resource-limited settings.

Automatic detection and segmentation of brain metastases on multimodal MR images with a deep convolutional neural network.

PubMed

Charron, Odelin; Lallement, Alex; Jarnet, Delphine; Noblet, Vincent; Clavier, Jean-Baptiste; Meyer, Philippe

2018-04-01

Stereotactic treatments are today the reference techniques for the irradiation of brain metastases in radiotherapy. The dose per fraction is very high, and delivered in small volumes (diameter <1 cm). As part of these treatments, effective detection and precise segmentation of lesions are imperative. Many methods based on deep-learning approaches have been developed for the automatic segmentation of gliomas, but very little for that of brain metastases. We adapted an existing 3D convolutional neural network (DeepMedic) to detect and segment brain metastases on MRI. At first, we sought to adapt the network parameters to brain metastases. We then explored the single or combined use of different MRI modalities, by evaluating network performance in terms of detection and segmentation. We also studied the interest of increasing the database with virtual patients or of using an additional database in which the active parts of the metastases are separated from the necrotic parts. Our results indicated that a deep network approach is promising for the detection and the segmentation of brain metastases on multimodal MRI. Copyright © 2018 Elsevier Ltd. All rights reserved.

Deep learning for brain tumor classification

NASA Astrophysics Data System (ADS)

Paul, Justin S.; Plassard, Andrew J.; Landman, Bennett A.; Fabbri, Daniel

2017-03-01

Recent research has shown that deep learning methods have performed well on supervised machine learning, image classification tasks. The purpose of this study is to apply deep learning methods to classify brain images with different tumor types: meningioma, glioma, and pituitary. A dataset was publicly released containing 3,064 T1-weighted contrast enhanced MRI (CE-MRI) brain images from 233 patients with either meningioma, glioma, or pituitary tumors split across axial, coronal, or sagittal planes. This research focuses on the 989 axial images from 191 patients in order to avoid confusing the neural networks with three different planes containing the same diagnosis. Two types of neural networks were used in classification: fully connected and convolutional neural networks. Within these two categories, further tests were computed via the augmentation of the original 512×512 axial images. Training neural networks over the axial data has proven to be accurate in its classifications with an average five-fold cross validation of 91.43% on the best trained neural network. This result demonstrates that a more general method (i.e. deep learning) can outperform specialized methods that require image dilation and ring-forming subregions on tumors.

Processing of chromatic information in a deep convolutional neural network.

PubMed

Flachot, Alban; Gegenfurtner, Karl R

2018-04-01

Deep convolutional neural networks are a class of machine-learning algorithms capable of solving non-trivial tasks, such as object recognition, with human-like performance. Little is known about the exact computations that deep neural networks learn, and to what extent these computations are similar to the ones performed by the primate brain. Here, we investigate how color information is processed in the different layers of the AlexNet deep neural network, originally trained on object classification of over 1.2M images of objects in their natural contexts. We found that the color-responsive units in the first layer of AlexNet learned linear features and were broadly tuned to two directions in color space, analogously to what is known of color responsive cells in the primate thalamus. Moreover, these directions are decorrelated and lead to statistically efficient representations, similar to the cardinal directions of the second-stage color mechanisms in primates. We also found, in analogy to the early stages of the primate visual system, that chromatic and achromatic information were segregated in the early layers of the network. Units in the higher layers of AlexNet exhibit on average a lower responsivity for color than units at earlier stages.

ABO blood groups and risk of deep venous thromboembolism in Chinese Han population from Chaoshan region in South China.

PubMed

Yu, Min; Wang, Cantian; Chen, Tingting; Hu, Shuang; Yi, Kaihong; Tan, Xuerui

2017-04-01

 Objectives: To demonstrate the prevalence of ABO blood groups with deep venous thromboembolism in Chinese Han population. A retrospective study was conducted between January 2010 and March 2015 in The First Affiliated Hospital of Shantou University Medical College in Chaoshan District of Guangdong Province in South China. Eighty nine patients with confirmed diagnosis of deep venous thromboembolism were included. Frequency of blood groups was determined. Results: Of 89 patients with deep venous thromboembolism, 28 patients had blood group A (31.5%), 28 patients had blood group B (31.5%), 13 patients had blood group AB (14.6%), and 20 patients had blood group O (22.5%). Compared with O blood type, the odds ratios of deep venous thromboembolism for A, B and AB were 2.23 (95% CI, 1.27-3.91), 2.34 (95% CI, 1.34-4.09) and  4.43 (95% CI, 2.24-8.76). Conclusion: There is a higher risk of venous thromboembolism in non-O blood groups than O group.

Archaeal Diversity in Waters from Deep South African Gold Mines

PubMed Central

Takai, Ken; Moser, Duane P.; DeFlaun, Mary; Onstott, Tullis C.; Fredrickson, James K.

2001-01-01

A culture-independent molecular analysis of archaeal communities in waters collected from deep South African gold mines was performed by performing a PCR-mediated terminal restriction fragment length polymorphism (T-RFLP) analysis of rRNA genes (rDNA) in conjunction with a sequencing analysis of archaeal rDNA clone libraries. The water samples used represented various environments, including deep fissure water, mine service water, and water from an overlying dolomite aquifer. T-RFLP analysis revealed that the ribotype distribution of archaea varied with the source of water. The archaeal communities in the deep gold mine environments exhibited great phylogenetic diversity; the majority of the members were most closely related to uncultivated species. Some archaeal rDNA clones obtained from mine service water and dolomite aquifer water samples were most closely related to environmental rDNA clones from surface soil (soil clones) and marine environments (marine group I [MGI]). Other clones exhibited intermediate phylogenetic affiliation between soil clones and MGI in the Crenarchaeota. Fissure water samples, derived from active or dormant geothermal environments, yielded archaeal sequences that exhibited novel phylogeny, including a novel lineage of Euryarchaeota. These results suggest that deep South African gold mines harbor novel archaeal communities distinct from those observed in other environments. Based on the phylogenetic analysis of archaeal strains and rDNA clones, including the newly discovered archaeal rDNA clones, the evolutionary relationship and the phylogenetic organization of the domain Archaea are reevaluated. PMID:11722932

Carbon and Neodymium Isotopic Fingerprints of Atlantic Deep Ocean Circulation During the Warm Pliocene

NASA Astrophysics Data System (ADS)

Riesselman, C. R.; Scher, H.; Robinson, M. M.; Dowsett, H. J.; Bell, D. B.

2012-12-01

Earth's future climate may resemble the mid-Piacenzian Age of the Pliocene, a time when global temperatures were sustained within the range predicted for the coming century. Surface and deep water temperature reconstructions and coupled ocean-atmosphere general circulation model simulations by the USGS PRISM (Pliocene Research Interpretation and Synoptic Mapping) Group identify a dramatic North Atlantic warm surface temperature anomaly in the mid-Piacenzian (3.264 - 3.025 Ma), accompanied by increased evaporation. The anomaly is detected in deep waters at 46°S, suggesting enhanced meridional overturning circulation and more southerly penetration of North Atlantic Deep Water (NADW) during the PRISM interval. However deep water temperature proxies are not diagnostic of water mass and some coupled model simulations predict transient decreases in NADW production in the 21st century, presenting a contrasting picture of future climate. We present a new multi-proxy investigation of Atlantic deep ocean circulation during the warm mid-Piacenzian, using δ13C of benthic foraminifera as a proxy for water mass age and the neodymium isotopic composition of fossil fish teeth (ɛNd) as a proxy for water mass source and mixing. This reconstruction utilizes both new and previously published data from DSDP and ODP cores along equatorial (Ceara Rise), southern mid-latitude (Walvis Ridge), and south Atlantic (Meteor Rise/Agulhas Ridge) depth transects. Additional end-member sites in the regions of modern north Atlantic and Southern Ocean deep water formation provide a Pliocene baseline for comparison. δ13C throughout the Atlantic basin is remarkably homogenous during the PRISM interval. δ13C values of Cibicidoides spp. and C. wuellerstorfi largely range between 0‰ and 1‰ at North Atlantic, shallow equatorial, southern mid-latitude, and south Atlantic sites with water depths from 2000-4700 m; both depth and latitudinal gradients are generally small (~0.3‰). However, equatorial Ceara Rise sites below 3500 m diverge, with δ13C values as low as -1.2‰ at ~3.15 Ma. The uniquely negative δ13C values at deep Ceara rise sites suggest that, during PRISM warmth, the oldest Atlantic deep waters may have resided along the modern deep western boundary current, while younger deep water masses were concentrated to the south and east. In the modern Atlantic, the ɛNd value of southern-sourced waters is more radiogenic than that of northern-sourced waters, providing a complimentary means to characterize Pliocene water mass geometry. ɛNd values from shallow (2500 m) and deep (4700 m) Walvis Ridge sites average -10 and -11 respectively; the shallow site is somewhat more radiogenic than published coretop ɛNd (-12), suggesting enhanced Pliocene influence of southern-sourced water masses. Ongoing analytical efforts will fingerprint Piacenzian ɛNd from north and south deep water source regions and will target additional depth transect ɛNd, allowing us to investigate the possibility that "older" carbon isotopic signatures at western equatorial sites reflect entrainment of proto-NADW while "younger" signatures at southern and eastern sites reflect the influence of southern-sourced deep water.

A novel end-to-end classifier using domain transferred deep convolutional neural networks for biomedical images.

PubMed

Pang, Shuchao; Yu, Zhezhou; Orgun, Mehmet A

2017-03-01

Highly accurate classification of biomedical images is an essential task in the clinical diagnosis of numerous medical diseases identified from those images. Traditional image classification methods combined with hand-crafted image feature descriptors and various classifiers are not able to effectively improve the accuracy rate and meet the high requirements of classification of biomedical images. The same also holds true for artificial neural network models directly trained with limited biomedical images used as training data or directly used as a black box to extract the deep features based on another distant dataset. In this study, we propose a highly reliable and accurate end-to-end classifier for all kinds of biomedical images via deep learning and transfer learning. We first apply domain transferred deep convolutional neural network for building a deep model; and then develop an overall deep learning architecture based on the raw pixels of original biomedical images using supervised training. In our model, we do not need the manual design of the feature space, seek an effective feature vector classifier or segment specific detection object and image patches, which are the main technological difficulties in the adoption of traditional image classification methods. Moreover, we do not need to be concerned with whether there are large training sets of annotated biomedical images, affordable parallel computing resources featuring GPUs or long times to wait for training a perfect deep model, which are the main problems to train deep neural networks for biomedical image classification as observed in recent works. With the utilization of a simple data augmentation method and fast convergence speed, our algorithm can achieve the best accuracy rate and outstanding classification ability for biomedical images. We have evaluated our classifier on several well-known public biomedical datasets and compared it with several state-of-the-art approaches. We propose a robust automated end-to-end classifier for biomedical images based on a domain transferred deep convolutional neural network model that shows a highly reliable and accurate performance which has been confirmed on several public biomedical image datasets. Copyright © 2017 Elsevier Ireland Ltd. All rights reserved.

Building the South Carolina Network for Educational Telecomputing. Four Years of Growth and Success--Bringing Educators Together. Annual Report to BellSouth Foundation. Project REACH, Telecommunications Grant (1 July 1992-1 April 1993).

ERIC Educational Resources Information Center

Oliver, S. Kemble, III

Expanding on its original purpose in developing an electronic community of humanities teachers and students in South Carolina from an existing agricultural telecommunications system, Project REACH received authorization from its sponsor, the BellSouth Foundation, to develop a statewide network for educational telecommunications and to provide…

Computations in the deep vs superficial layers of the cerebral cortex.

PubMed

Rolls, Edmund T; Mills, W Patrick C

2017-11-01

A fundamental question is how the cerebral neocortex operates functionally, computationally. The cerebral neocortex with its superficial and deep layers and highly developed recurrent collateral systems that provide a basis for memory-related processing might perform somewhat different computations in the superficial and deep layers. Here we take into account the quantitative connectivity within and between laminae. Using integrate-and-fire neuronal network simulations that incorporate this connectivity, we first show that attractor networks implemented in the deep layers that are activated by the superficial layers could be partly independent in that the deep layers might have a different time course, which might because of adaptation be more transient and useful for outputs from the neocortex. In contrast the superficial layers could implement more prolonged firing, useful for slow learning and for short-term memory. Second, we show that a different type of computation could in principle be performed in the superficial and deep layers, by showing that the superficial layers could operate as a discrete attractor network useful for categorisation and feeding information forward up a cortical hierarchy, whereas the deep layers could operate as a continuous attractor network useful for providing a spatially and temporally smooth output to output systems in the brain. A key advance is that we draw attention to the functions of the recurrent collateral connections between cortical pyramidal cells, often omitted in canonical models of the neocortex, and address principles of operation of the neocortex by which the superficial and deep layers might be specialized for different types of attractor-related memory functions implemented by the recurrent collaterals. Copyright © 2017 Elsevier Inc. All rights reserved.

Telemedicine in Western Africa: lessons learned from a pilot project in Mali, perspectives and recommendations.

PubMed

Geissbuhler, Antoine; Ly, Ousmane; Lovis, Christian; L'Haire, Jean-François

2003-01-01

to evaluate the feasibility, potential and risks of an internet-based telemedicine network in developing countries of Western Africa. a project for the development of a national telemedicine network in Mali was initiated in 2001, using internet-based technologies for distance learning and teleconsultations. the telemedicine network has been in productive use for 12 months and has enabled various collaboration channels, including North-South, South-South, and South-North distance learning and teleconsultations. It also unveiled a set of potential problems: a) limited pertinence of North-South collaborations when there are major differences in available resources or socio-cultural contexts between the collaborating parties; b) risk of induced digital divide if the periphery of the health system is not involved in the development of the network, and c) need for the development of local medical contents management skills. the identified risks must be taken into account when designing large-scale telemedicine projects in developing countries and can be mitigated by the fostering of South-South collaboration channels, the use of satellite-based internet connectivity in remote areas, and the valorization of local knowledge and its publication on-line.

Establishing a Multi-scale Stream Gaging Network in the Whitewater River Basin, Kansas, USA

USGS Publications Warehouse

Clayton, J.A.; Kean, J.W.

2010-01-01

Investigating the routing of streamflow through a large drainage basin requires the determination of discharge at numerous locations in the channel network. Establishing a dense network of stream gages using conventional methods is both cost-prohibitive and functionally impractical for many research projects. We employ herein a previously tested, fluid-mechanically based model for generating rating curves to establish a stream gaging network in the Whitewater River basin in south-central Kansas. The model was developed for the type of channels typically found in this watershed, meaning that it is designed to handle deep, narrow geomorphically stable channels with irregular planforms, and can model overbank flow over a vegetated floodplain. We applied the model to ten previously ungaged stream reaches in the basin, ranging from third- to sixth-order channels. At each site, detailed field measurements of the channel and floodplain morphology, bed and bank roughness, and vegetation characteristics were used to quantify the roughness for a range of flow stages, from low flow to overbank flooding. Rating curves that relate stage to discharge were developed for all ten sites. Both fieldwork and modeling were completed in less than 2 years during an anomalously dry period in the region, which underscores an advantage of using theoretically based (as opposed to empirically based) discharge estimation techniques. ?? 2010 Springer Science+Business Media B.V.

Adaptive template generation for amyloid PET using a deep learning approach.

PubMed

Kang, Seung Kwan; Seo, Seongho; Shin, Seong A; Byun, Min Soo; Lee, Dong Young; Kim, Yu Kyeong; Lee, Dong Soo; Lee, Jae Sung

2018-05-11

Accurate spatial normalization (SN) of amyloid positron emission tomography (PET) images for Alzheimer's disease assessment without coregistered anatomical magnetic resonance imaging (MRI) of the same individual is technically challenging. In this study, we applied deep neural networks to generate individually adaptive PET templates for robust and accurate SN of amyloid PET without using matched 3D MR images. Using 681 pairs of simultaneously acquired 11 C-PIB PET and T1-weighted 3D MRI scans of AD, MCI, and cognitively normal subjects, we trained and tested two deep neural networks [convolutional auto-encoder (CAE) and generative adversarial network (GAN)] that produce adaptive best PET templates. More specifically, the networks were trained using 685,100 pieces of augmented data generated by rotating 527 randomly selected datasets and validated using 154 datasets. The input to the supervised neural networks was the 3D PET volume in native space and the label was the spatially normalized 3D PET image using the transformation parameters obtained from MRI-based SN. The proposed deep learning approach significantly enhanced the quantitative accuracy of MRI-less amyloid PET assessment by reducing the SN error observed when an average amyloid PET template is used. Given an input image, the trained deep neural networks rapidly provide individually adaptive 3D PET templates without any discontinuity between the slices (in 0.02 s). As the proposed method does not require 3D MRI for the SN of PET images, it has great potential for use in routine analysis of amyloid PET images in clinical practice and research. © 2018 Wiley Periodicals, Inc.

Physical activity levels and perceived benefits and barriers to physical activity in HIV-infected women living in the deep south of the United States.

PubMed

Rehm, Kristina E; Konkle-Parker, Deborah

2016-09-01

Engaging in regular physical activity (PA) is important in maintaining health and increasing the overall quality of life of people living with HIV (PLWH). The deep south of the USA is known for its high rate of sedentary behavior although data on the activity levels and perceptions of the benefits and barriers to exercise in women living with HIV in the deep south are lacking. Understanding the perceived benefits and barriers to exercise can guide the development of PA interventions. We conducted a cross-sectional study to determine the PA levels and perceived benefits and barriers to exercise associated with both age and depression level in a group of HIV+ women living in the deep south. We recruited a total of 50 participants from a cohort site for the Women's Interagency HIV Study. Depression was assessed using the Center for Epidemiological Studies Depression Scale (CES-D) and benefits/barriers to exercise were measured using the Exercise Benefits and Barriers Scale (EBBS). We measured PA both subjectively and objectively using the International Physical Activity Questionnaire (IPAQ) and a Fitbit PA monitor, respectively. Our sample was predominantly African-American (96%) and the mean ±SD age, body mass index, and CES-D score were 42 ± 8.8 years, 36.6 ± 11.5 kg/m(2), and 15.6 ± 11.4, respectively. Both subjective and objective measures of PA indicated that our participants were sedentary. The greatest perceived benefit to exercise was physical performance and the greatest barrier to exercise was physical exertion. Higher overall perceived benefits were reported by women ≥43 years and women reporting higher levels of depression. There was no difference in overall barriers associated with age and depression level, but women with depression felt more fatigued by exercise. The results of this study can be helpful when designing and implementing PA interventions in women living with HIV in the deep south.

A deep convolutional neural network to analyze position averaged convergent beam electron diffraction patterns.

PubMed

Xu, W; LeBeau, J M

2018-05-01

We establish a series of deep convolutional neural networks to automatically analyze position averaged convergent beam electron diffraction patterns. The networks first calibrate the zero-order disk size, center position, and rotation without the need for pretreating the data. With the aligned data, additional networks then measure the sample thickness and tilt. The performance of the network is explored as a function of a variety of variables including thickness, tilt, and dose. A methodology to explore the response of the neural network to various pattern features is also presented. Processing patterns at a rate of  ∼ 0.1 s/pattern, the network is shown to be orders of magnitude faster than a brute force method while maintaining accuracy. The approach is thus suitable for automatically processing big, 4D STEM data. We also discuss the generality of the method to other materials/orientations as well as a hybrid approach that combines the features of the neural network with least squares fitting for even more robust analysis. The source code is available at https://github.com/subangstrom/DeepDiffraction. Copyright © 2018 Elsevier B.V. All rights reserved.

Deep Learning and Its Applications in Biomedicine.

PubMed

Cao, Chensi; Liu, Feng; Tan, Hai; Song, Deshou; Shu, Wenjie; Li, Weizhong; Zhou, Yiming; Bo, Xiaochen; Xie, Zhi

2018-02-01

Advances in biological and medical technologies have been providing us explosive volumes of biological and physiological data, such as medical images, electroencephalography, genomic and protein sequences. Learning from these data facilitates the understanding of human health and disease. Developed from artificial neural networks, deep learning-based algorithms show great promise in extracting features and learning patterns from complex data. The aim of this paper is to provide an overview of deep learning techniques and some of the state-of-the-art applications in the biomedical field. We first introduce the development of artificial neural network and deep learning. We then describe two main components of deep learning, i.e., deep learning architectures and model optimization. Subsequently, some examples are demonstrated for deep learning applications, including medical image classification, genomic sequence analysis, as well as protein structure classification and prediction. Finally, we offer our perspectives for the future directions in the field of deep learning. Copyright © 2018. Production and hosting by Elsevier B.V.

Operability engineering in the Deep Space Network

NASA Technical Reports Server (NTRS)

Wilkinson, Belinda

1993-01-01

Many operability problems exist at the three Deep Space Communications Complexes (DSCC's) of the Deep Space Network (DSN). Four years ago, the position of DSN Operability Engineer was created to provide the opportunity for someone to take a system-level approach to solving these problems. Since that time, a process has been developed for personnel and development engineers and for enforcing user interface standards in software designed for the DSCC's. Plans are for the participation of operations personnel in the product life-cycle to expand in the future.

Deep Potential Molecular Dynamics: A Scalable Model with the Accuracy of Quantum Mechanics

NASA Astrophysics Data System (ADS)

Zhang, Linfeng; Han, Jiequn; Wang, Han; Car, Roberto; E, Weinan

2018-04-01

We introduce a scheme for molecular simulations, the deep potential molecular dynamics (DPMD) method, based on a many-body potential and interatomic forces generated by a carefully crafted deep neural network trained with ab initio data. The neural network model preserves all the natural symmetries in the problem. It is first-principles based in the sense that there are no ad hoc components aside from the network model. We show that the proposed scheme provides an efficient and accurate protocol in a variety of systems, including bulk materials and molecules. In all these cases, DPMD gives results that are essentially indistinguishable from the original data, at a cost that scales linearly with system size.

[Scientific production in clinical medicine and international collaboration networks in South American countries].

PubMed

Huamaní, Charles; González A, Gregorio; Curioso, Walter H; Pacheco-Romero, José

2012-04-01

International collaboration is increasingly used in biomedical research. To describe the characteristics of scientific production in Latin America and the main international collaboration networks for the period 2000 to 2009. Search for papers generated in Latin American countries in the Clinical Medicine database of ISI Web of Knowledge v.4.10 - Current Contents Connect. The country of origin of the corresponding author was considered the producing country of the paper. International collaboration was analyzed calculating the number of countries that contributed to the generation of a particular paper. Collaboration networks were graphed to determine the centrality of each network. Twelve Latin American countries participated in the production of 253,362 papers. The corresponding author was South American in 79% of these papers. Sixteen percent of papers were on clinical medicine and 36% of these were carried out in collaboration. Brazil had the highest production (22,442 papers) and the lower percentage of international collaboration (31%). North America accounts for 63% of collaborating countries. Only 8% of collaboration is between South American countries. Brazil has the highest tendency to collaborate with other South American countries. Brazil is the South American country with the highest scientific production and indicators of centrality in South America. The most common collaboration networks are with North American countries.

Using Deep Learning for Gamma Ray Source Detection at the First G-APD Cherenkov Telescope (FACT)

NASA Astrophysics Data System (ADS)

Bieker, Jacob

2018-06-01

Finding gamma-ray sources is of paramount importance for Imaging Air Cherenkov Telescopes (IACT). This study looks at using deep neural networks on data from the First G-APD Cherenkov Telescope (FACT) as a proof-of-concept of finding gamma-ray sources with deep learning for the upcoming Cherenkov Telescope Array (CTA). In this study, FACT’s individual photon level observation data from the last 5 years was used with convolutional neural networks to determine if one or more sources were present. The neural networks used various architectures to determine which architectures were most successful in finding sources. Neural networks offer a promising method for finding faint and extended gamma-ray sources for IACTs. With further improvement and modifications, they offer a compelling method for source detection for the next generation of IACTs.

Pulmonary Nodule Classification with Deep Convolutional Neural Networks on Computed Tomography Images.

PubMed

Li, Wei; Cao, Peng; Zhao, Dazhe; Wang, Junbo

2016-01-01

Computer aided detection (CAD) systems can assist radiologists by offering a second opinion on early diagnosis of lung cancer. Classification and feature representation play critical roles in false-positive reduction (FPR) in lung nodule CAD. We design a deep convolutional neural networks method for nodule classification, which has an advantage of autolearning representation and strong generalization ability. A specified network structure for nodule images is proposed to solve the recognition of three types of nodules, that is, solid, semisolid, and ground glass opacity (GGO). Deep convolutional neural networks are trained by 62,492 regions-of-interest (ROIs) samples including 40,772 nodules and 21,720 nonnodules from the Lung Image Database Consortium (LIDC) database. Experimental results demonstrate the effectiveness of the proposed method in terms of sensitivity and overall accuracy and that it consistently outperforms the competing methods.

Quantifying Potential Groundwater Recharge In South Texas

NASA Astrophysics Data System (ADS)

Basant, S.; Zhou, Y.; Leite, P. A.; Wilcox, B. P.

2015-12-01

Groundwater in South Texas is heavily relied on for human consumption and irrigation for food crops. Like most of the south west US, woody encroachment has altered the grassland ecosystems here too. While brush removal has been widely implemented in Texas with the objective of increasing groundwater recharge, the linkage between vegetation and groundwater recharge in South Texas is still unclear. Studies have been conducted to understand plant-root-water dynamics at the scale of plants. However, little work has been done to quantify the changes in soil water and deep percolation at the landscape scale. Modeling water flow through soil profiles can provide an estimate of the total water flowing into deep percolation. These models are especially powerful with parameterized and calibrated with long term soil water data. In this study we parameterize the HYDRUS soil water model using long term soil water data collected in Jim Wells County in South Texas. Soil water was measured at every 20 cm intervals up to a depth of 200 cm. The parameterized model will be used to simulate soil water dynamics under a variety of precipitation regimes ranging from well above normal to severe drought conditions. The results from the model will be compared with the changes in soil moisture profile observed in response to vegetation cover and treatments from a study in a similar. Comparative studies like this can be used to build new and strengthen existing hypotheses regarding deep percolation and the role of soil texture and vegetation in groundwater recharge.

Intense Mixing and Recirculations of Intermediate and Deep Water in the Northwest Argentine Basin

NASA Astrophysics Data System (ADS)

Valla, D.; Piola, A. R.

2016-02-01

The sources of the South Atlantic upper and intermediate waters that form the upper layer flow needed to maintain mass balance due the export of North Atlantic Deep Water from the North Atlantic are still under debate. The "cold path" scheme postulates that intermediate waters are injected into the South Atlantic from the Pacific through the Drake Passage, advected north by the Malvinas Current up to the Brazil/Malvinas Confluence (BMC) and circulated around the basin following the path of the subtropical gyre. We report high-quality hydrographic observations collected in the South Atlantic western boundary at 34.5 °S during 7 hydrographic cruises as part of the SAMOC project. We focus on the flow and characteristics of Antarctic Intermediate Water (AAIW) and Upper Circumpolar Deep Water (UCDW). The water mass analysis indicates the presence of "young" (fresh and highly oxygenated) varieties of AAIW (S<34.2, O2>6 ml·l-1) which must be derived from south of the SAMOC array. This suggests an alternative pathway for intermediate waters that involves a short circuit beneath the BMC. Simultaneous full-depth velocity measurements using lowered acoustic Doppler current profilers confirm this hypothesis. The flow direction across the SAMOC array in the UCDW range inferred from dissolved oxygen measurements also indicate the presence of UCDW (O2<4.2 ml·l-1) derived from farther south. However, the wider range of oxygen concentrations suggests strong recirculations of both water masses within the northwestern Argentine Basin.

Quantifying the Digital Divide: A Scientific Overview of Network Connectivity and Grid Infrastructure in South Asian Countries

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

Khan, Shahryar Muhammad; /SLAC /NUST, Rawalpindi; Cottrell, R.Les

2007-10-30

The future of Computing in High Energy Physics (HEP) applications depends on both the Network and Grid infrastructure. South Asian countries such as India and Pakistan are making significant progress by building clusters as well as improving their network infrastructure However to facilitate the use of these resources, they need to manage the issues of network connectivity to be among the leading participants in Computing for HEP experiments. In this paper we classify the connectivity for academic and research institutions of South Asia. The quantitative measurements are carried out using the PingER methodology; an approach that induces minimal ICMP trafficmore » to gather active end-to-end network statistics. The PingER project has been measuring the Internet performance for the last decade. Currently the measurement infrastructure comprises of over 700 hosts in more than 130 countries which collectively represents approximately 99% of the world's Internet-connected population. Thus, we are well positioned to characterize the world's connectivity. Here we present the current state of the National Research and Educational Networks (NRENs) and Grid Infrastructure in the South Asian countries and identify the areas of concern. We also present comparisons between South Asia and other developing as well as developed regions. We show that there is a strong correlation between the Network performance and several Human Development indices.« less

Nitrogen isotopic composition of nitrate in the South China Sea: A clue to the origin of nitrogen

NASA Astrophysics Data System (ADS)

Yang, Z.; Chen, J.; Chen, M.; Ran, L.; Li, H.; Zhu, Y.; Wang, C.; Ji, Z.; Zhang, J.; Zhang, D.

2016-02-01

Nitrogen isotopic composition of water column nitrate was measured in the South China Sea to clarify the sources of nitrogen. The δ15NNO3 value in deep water (5.4±0.2‰) was higher than the average deep oceanic δ15NNO3 ( 5‰), and a weak δ15NNO3 maximum (5.9±0.2‰) was observed at 500 m depth, matching the salinity minimum. These indicated the intrusion of the North Pacific Water which carried nitrate with a high δ15NNO3 and showed a similar δ15NNO3 distribution profile with the South China Sea. The high N* (1.74±0.23 μmol/L) combined with the low δ15NNO3 (4.7±0.2‰) at 100 m depth indicated that N2 fixation (and possibly Atmospheric Deposition) introduces new N to the surface ocean. The distribution of δ15N values of nitrate, sinking particles and surface sediment suggest that laterally-advected sediments may be a source of nitrogen to the deep ocean.

Explaining and Improving Breast Cancer Information Acquisition among African American Women in the Deep South

PubMed Central

Anderson-Lewis, Charkarra; Ross, Levi; Johnson, Jarrett; Hastrup, Janice L.; Green, B. Lee; Kohler, Connie L.

2012-01-01

Objectives A major challenge facing contemporary cancer educators is how to optimize the dissemination of breast cancer prevention and control information to African American women in the Deep South who are believed to be cancer free. The purpose of this research was to provide insight into the breast cancer information-acquisition experiences of African American women in Alabama and Mississippi and to make recommendations on ways to better reach members of this high-risk, underserved population. Methods Focus group methodology was used in a repeated, cross-sectional research design with 64 African American women, 35 years old or older who lived in one of four urban or rural counties in Alabama and Mississippi. Results Axial-coded themes emerged around sources of cancer information, patterns of information acquisition, characteristics of preferred sources, and characteristics of least-preferred sources. Conclusions It is important to invest in lay health educators to optimize the dissemination of breast cancer information to African American women who are believed to be cancer free in the Deep South. PMID:22665151

Deep Gaze Velocity Analysis During Mammographic Reading for Biometric Identification of Radiologists

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

Yoon, Hong-Jun; Alamudun, Folami T.; Hudson, Kathy

Several studies have confirmed that the gaze velocity of the human eye can be utilized as a behavioral biometric or personalized biomarker. In this study, we leverage the local feature representation capacity of convolutional neural networks (CNNs) for eye gaze velocity analysis as the basis for biometric identification of radiologists performing breast cancer screening. Using gaze data collected from 10 radiologists reading 100 mammograms of various diagnoses, we compared the performance of a CNN-based classification algorithm with two deep learning classifiers, deep neural network and deep belief network, and a previously presented hidden Markov model classifier. The study showed thatmore » the CNN classifier is superior compared to alternative classification methods based on macro F1-scores derived from 10-fold cross-validation experiments. Our results further support the efficacy of eye gaze velocity as a biometric identifier of medical imaging experts.« less

Deep Gaze Velocity Analysis During Mammographic Reading for Biometric Identification of Radiologists

DOE PAGES

Yoon, Hong-Jun; Alamudun, Folami T.; Hudson, Kathy; ...

2018-01-24

Several studies have confirmed that the gaze velocity of the human eye can be utilized as a behavioral biometric or personalized biomarker. In this study, we leverage the local feature representation capacity of convolutional neural networks (CNNs) for eye gaze velocity analysis as the basis for biometric identification of radiologists performing breast cancer screening. Using gaze data collected from 10 radiologists reading 100 mammograms of various diagnoses, we compared the performance of a CNN-based classification algorithm with two deep learning classifiers, deep neural network and deep belief network, and a previously presented hidden Markov model classifier. The study showed thatmore » the CNN classifier is superior compared to alternative classification methods based on macro F1-scores derived from 10-fold cross-validation experiments. Our results further support the efficacy of eye gaze velocity as a biometric identifier of medical imaging experts.« less

Enhanced Higgs boson to τ(+)τ(-) search with deep learning.

PubMed

Baldi, P; Sadowski, P; Whiteson, D

2015-03-20

The Higgs boson is thought to provide the interaction that imparts mass to the fundamental fermions, but while measurements at the Large Hadron Collider (LHC) are consistent with this hypothesis, current analysis techniques lack the statistical power to cross the traditional 5σ significance barrier without more data. Deep learning techniques have the potential to increase the statistical power of this analysis by automatically learning complex, high-level data representations. In this work, deep neural networks are used to detect the decay of the Higgs boson to a pair of tau leptons. A Bayesian optimization algorithm is used to tune the network architecture and training algorithm hyperparameters, resulting in a deep network of eight nonlinear processing layers that improves upon the performance of shallow classifiers even without the use of features specifically engineered by physicists for this application. The improvement in discovery significance is equivalent to an increase in the accumulated data set of 25%.

Prediction of residue-residue contact matrix for protein-protein interaction with Fisher score features and deep learning.

PubMed

Du, Tianchuan; Liao, Li; Wu, Cathy H; Sun, Bilin

2016-11-01

Protein-protein interactions play essential roles in many biological processes. Acquiring knowledge of the residue-residue contact information of two interacting proteins is not only helpful in annotating functions for proteins, but also critical for structure-based drug design. The prediction of the protein residue-residue contact matrix of the interfacial regions is challenging. In this work, we introduced deep learning techniques (specifically, stacked autoencoders) to build deep neural network models to tackled the residue-residue contact prediction problem. In tandem with interaction profile Hidden Markov Models, which was used first to extract Fisher score features from protein sequences, stacked autoencoders were deployed to extract and learn hidden abstract features. The deep learning model showed significant improvement over the traditional machine learning model, Support Vector Machines (SVM), with the overall accuracy increased by 15% from 65.40% to 80.82%. We showed that the stacked autoencoders could extract novel features, which can be utilized by deep neural networks and other classifiers to enhance learning, out of the Fisher score features. It is further shown that deep neural networks have significant advantages over SVM in making use of the newly extracted features. Copyright © 2016. Published by Elsevier Inc.

Underwater Inherent Optical Properties Estimation Using a Depth Aided Deep Neural Network.

PubMed

Yu, Zhibin; Wang, Yubo; Zheng, Bing; Zheng, Haiyong; Wang, Nan; Gu, Zhaorui

2017-01-01

Underwater inherent optical properties (IOPs) are the fundamental clues to many research fields such as marine optics, marine biology, and underwater vision. Currently, beam transmissometers and optical sensors are considered as the ideal IOPs measuring methods. But these methods are inflexible and expensive to be deployed. To overcome this problem, we aim to develop a novel measuring method using only a single underwater image with the help of deep artificial neural network. The power of artificial neural network has been proved in image processing and computer vision fields with deep learning technology. However, image-based IOPs estimation is a quite different and challenging task. Unlike the traditional applications such as image classification or localization, IOP estimation looks at the transparency of the water between the camera and the target objects to estimate multiple optical properties simultaneously. In this paper, we propose a novel Depth Aided (DA) deep neural network structure for IOPs estimation based on a single RGB image that is even noisy. The imaging depth information is considered as an aided input to help our model make better decision.

Preliminary Concept of Operations for the Deep Space Array-Based Network

NASA Astrophysics Data System (ADS)

Bagri, D. S.; Statman, J. I.

2004-05-01

The Deep Space Array-Based Network (DSAN) will be an array-based system, part of a greater than 1000 times increase in the downlink/telemetry capability of the Deep Space Network. The key function of the DSAN is provision of cost-effective, robust telemetry, tracking, and command services to the space missions of NASA and its international partners. This article presents an expanded approach to the use of an array-based system. Instead of using the array as an element in the existing Deep Space Network (DSN), relying to a large extent on the DSN infrastructure, we explore a broader departure from the current DSN, using fewer elements of the existing DSN, and establishing a more modern concept of operations. For example, the DSAN will have a single 24 x 7 monitor and control (M&C) facility, while the DSN has four 24 x 7 M&C facilities. The article gives the architecture of the DSAN and its operations philosophy. It also briefly describes the customer's view of operations, operations management, logistics, anomaly analysis, and reporting.

TopologyNet: Topology based deep convolutional and multi-task neural networks for biomolecular property predictions

PubMed Central

2017-01-01

Although deep learning approaches have had tremendous success in image, video and audio processing, computer vision, and speech recognition, their applications to three-dimensional (3D) biomolecular structural data sets have been hindered by the geometric and biological complexity. To address this problem we introduce the element-specific persistent homology (ESPH) method. ESPH represents 3D complex geometry by one-dimensional (1D) topological invariants and retains important biological information via a multichannel image-like representation. This representation reveals hidden structure-function relationships in biomolecules. We further integrate ESPH and deep convolutional neural networks to construct a multichannel topological neural network (TopologyNet) for the predictions of protein-ligand binding affinities and protein stability changes upon mutation. To overcome the deep learning limitations from small and noisy training sets, we propose a multi-task multichannel topological convolutional neural network (MM-TCNN). We demonstrate that TopologyNet outperforms the latest methods in the prediction of protein-ligand binding affinities, mutation induced globular protein folding free energy changes, and mutation induced membrane protein folding free energy changes. Availability: weilab.math.msu.edu/TDL/ PMID:28749969

Deep Recurrent Neural Networks for Human Activity Recognition

PubMed Central

Murad, Abdulmajid

2017-01-01

Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs. PMID:29113103

Deep Recurrent Neural Networks for Human Activity Recognition.

PubMed

Murad, Abdulmajid; Pyun, Jae-Young

2017-11-06

Adopting deep learning methods for human activity recognition has been effective in extracting discriminative features from raw input sequences acquired from body-worn sensors. Although human movements are encoded in a sequence of successive samples in time, typical machine learning methods perform recognition tasks without exploiting the temporal correlations between input data samples. Convolutional neural networks (CNNs) address this issue by using convolutions across a one-dimensional temporal sequence to capture dependencies among input data. However, the size of convolutional kernels restricts the captured range of dependencies between data samples. As a result, typical models are unadaptable to a wide range of activity-recognition configurations and require fixed-length input windows. In this paper, we propose the use of deep recurrent neural networks (DRNNs) for building recognition models that are capable of capturing long-range dependencies in variable-length input sequences. We present unidirectional, bidirectional, and cascaded architectures based on long short-term memory (LSTM) DRNNs and evaluate their effectiveness on miscellaneous benchmark datasets. Experimental results show that our proposed models outperform methods employing conventional machine learning, such as support vector machine (SVM) and k-nearest neighbors (KNN). Additionally, the proposed models yield better performance than other deep learning techniques, such as deep believe networks (DBNs) and CNNs.

LiteNet: Lightweight Neural Network for Detecting Arrhythmias at Resource-Constrained Mobile Devices.

PubMed

He, Ziyang; Zhang, Xiaoqing; Cao, Yangjie; Liu, Zhi; Zhang, Bo; Wang, Xiaoyan

2018-04-17

By running applications and services closer to the user, edge processing provides many advantages, such as short response time and reduced network traffic. Deep-learning based algorithms provide significantly better performances than traditional algorithms in many fields but demand more resources, such as higher computational power and more memory. Hence, designing deep learning algorithms that are more suitable for resource-constrained mobile devices is vital. In this paper, we build a lightweight neural network, termed LiteNet which uses a deep learning algorithm design to diagnose arrhythmias, as an example to show how we design deep learning schemes for resource-constrained mobile devices. Compare to other deep learning models with an equivalent accuracy, LiteNet has several advantages. It requires less memory, incurs lower computational cost, and is more feasible for deployment on resource-constrained mobile devices. It can be trained faster than other neural network algorithms and requires less communication across different processing units during distributed training. It uses filters of heterogeneous size in a convolutional layer, which contributes to the generation of various feature maps. The algorithm was tested using the MIT-BIH electrocardiogram (ECG) arrhythmia database; the results showed that LiteNet outperforms comparable schemes in diagnosing arrhythmias, and in its feasibility for use at the mobile devices.

LiteNet: Lightweight Neural Network for Detecting Arrhythmias at Resource-Constrained Mobile Devices

PubMed Central

Zhang, Xiaoqing; Cao, Yangjie; Liu, Zhi; Zhang, Bo; Wang, Xiaoyan

2018-01-01

By running applications and services closer to the user, edge processing provides many advantages, such as short response time and reduced network traffic. Deep-learning based algorithms provide significantly better performances than traditional algorithms in many fields but demand more resources, such as higher computational power and more memory. Hence, designing deep learning algorithms that are more suitable for resource-constrained mobile devices is vital. In this paper, we build a lightweight neural network, termed LiteNet which uses a deep learning algorithm design to diagnose arrhythmias, as an example to show how we design deep learning schemes for resource-constrained mobile devices. Compare to other deep learning models with an equivalent accuracy, LiteNet has several advantages. It requires less memory, incurs lower computational cost, and is more feasible for deployment on resource-constrained mobile devices. It can be trained faster than other neural network algorithms and requires less communication across different processing units during distributed training. It uses filters of heterogeneous size in a convolutional layer, which contributes to the generation of various feature maps. The algorithm was tested using the MIT-BIH electrocardiogram (ECG) arrhythmia database; the results showed that LiteNet outperforms comparable schemes in diagnosing arrhythmias, and in its feasibility for use at the mobile devices. PMID:29673171

An introduction to deep learning on biological sequence data: examples and solutions.

PubMed

Jurtz, Vanessa Isabell; Johansen, Alexander Rosenberg; Nielsen, Morten; Almagro Armenteros, Jose Juan; Nielsen, Henrik; Sønderby, Casper Kaae; Winther, Ole; Sønderby, Søren Kaae

2017-11-15

Deep neural network architectures such as convolutional and long short-term memory networks have become increasingly popular as machine learning tools during the recent years. The availability of greater computational resources, more data, new algorithms for training deep models and easy to use libraries for implementation and training of neural networks are the drivers of this development. The use of deep learning has been especially successful in image recognition; and the development of tools, applications and code examples are in most cases centered within this field rather than within biology. Here, we aim to further the development of deep learning methods within biology by providing application examples and ready to apply and adapt code templates. Given such examples, we illustrate how architectures consisting of convolutional and long short-term memory neural networks can relatively easily be designed and trained to state-of-the-art performance on three biological sequence problems: prediction of subcellular localization, protein secondary structure and the binding of peptides to MHC Class II molecules. All implementations and datasets are available online to the scientific community at https://github.com/vanessajurtz/lasagne4bio. skaaesonderby@gmail.com. Supplementary data are available at Bioinformatics online. © The Author (2017). Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Deep Space Telecommunications

NASA Technical Reports Server (NTRS)

Kuiper, T. B. H.; Resch, G. M.

2000-01-01

The increasing load on NASA's deep Space Network, the new capabilities for deep space missions inherent in a next-generation radio telescope, and the potential of new telescope technology for reducing construction and operation costs suggest a natural marriage between radio astronomy and deep space telecommunications in developing advanced radio telescope concepts.

The EMBRACE Magnetometer Network in South America: Network Description and Firsts Results

NASA Astrophysics Data System (ADS)

Denardini, Clezio Marcos

We present the new EMBRACE Magnetometer Network in South America, which so far is planned to cover most of the Easter Southern American longitudinal sector deploying magnetometer in several locations. We discuss the purpose and scientific goals of the network, associated with the Low- and Mid-Latitude Ionospheric Currents and Space Weather. We provide details on the instrumentation, the inter-calibration procedure, and installations of equipments already installed. In addition, we present and discuss details on the data storage, near-real time display and availability. Finally, we provide some of the first results we already achieved from this network, including the development of new real time magnetic regional indices for South America. Contacting Author: C. M. Denardini (clezio.denardin@inpe.br)

Decade-long deep-ocean warming detected in the subtropical South Pacific

PubMed Central

Volkov, Denis L.; Lee, Sang-Ki; Landerer, Felix W.; Lumpkin, Rick

2017-01-01

The persistent energy imbalance at the top of the atmosphere, inferred from satellite measurements, indicates that the Earth’s climate system continues to accumulate excess heat. As only sparse and irregular measurements of ocean heat below 2000 m depth exist, one of the most challenging questions in global climate change studies is whether the excess heat has already penetrated into the deep ocean. Here we perform a comprehensive analysis of satellite and in situ measurements to report that a significant deep-ocean warming occurred in the subtropical South Pacific Ocean over the past decade (2005–2014). The local accumulation of heat accounted for up to a quarter of the global ocean heat increase, with directly and indirectly inferred deep ocean (below 2000 m) contribution of 2.4 ± 1.4 and 6.1–10.1 ± 4.4%, respectively. We further demonstrate that this heat accumulation is consistent with a decade-long intensification of the subtropical convergence, possibly linked to the persistent La Niña-like state. PMID:29200536

Decade-long deep-ocean warming detected in the subtropical South Pacific.

PubMed

Volkov, Denis L; Lee, Sang-Ki; Landerer, Felix W; Lumpkin, Rick

2017-01-28

The persistent energy imbalance at the top of the atmosphere, inferred from satellite measurements, indicates that the Earth's climate system continues to accumulate excess heat. As only sparse and irregular measurements of ocean heat below 2000 m depth exist, one of the most challenging questions in global climate change studies is whether the excess heat has already penetrated into the deep ocean. Here we perform a comprehensive analysis of satellite and in situ measurements to report that a significant deep-ocean warming occurred in the subtropical South Pacific Ocean over the past decade (2005-2014). The local accumulation of heat accounted for up to a quarter of the global ocean heat increase, with directly and indirectly inferred deep ocean (below 2000 m) contribution of 2.4 ± 1.4 and 6.1-10.1 ± 4.4%, respectively. We further demonstrate that this heat accumulation is consistent with a decade-long intensification of the subtropical convergence, possibly linked to the persistent La Niña-like state.

Deep Neural Network for Structural Prediction and Lane Detection in Traffic Scene.

PubMed

Li, Jun; Mei, Xue; Prokhorov, Danil; Tao, Dacheng

2017-03-01

Hierarchical neural networks have been shown to be effective in learning representative image features and recognizing object classes. However, most existing networks combine the low/middle level cues for classification without accounting for any spatial structures. For applications such as understanding a scene, how the visual cues are spatially distributed in an image becomes essential for successful analysis. This paper extends the framework of deep neural networks by accounting for the structural cues in the visual signals. In particular, two kinds of neural networks have been proposed. First, we develop a multitask deep convolutional network, which simultaneously detects the presence of the target and the geometric attributes (location and orientation) of the target with respect to the region of interest. Second, a recurrent neuron layer is adopted for structured visual detection. The recurrent neurons can deal with the spatial distribution of visible cues belonging to an object whose shape or structure is difficult to explicitly define. Both the networks are demonstrated by the practical task of detecting lane boundaries in traffic scenes. The multitask convolutional neural network provides auxiliary geometric information to help the subsequent modeling of the given lane structures. The recurrent neural network automatically detects lane boundaries, including those areas containing no marks, without any explicit prior knowledge or secondary modeling.

Mantle Upwellings Below the Ibero-Maghrebian Region with a Common Deep Source from P Travel-time Tomography

NASA Astrophysics Data System (ADS)

Civiero, C.; Custodio, S.; Silveira, G. M.; Rawlinson, N.; Arroucau, P.

2017-12-01

The processes responsible for the geodynamical evolution of the Ibero-Maghrebian domain are still enigmatic. Several geophysical studies have improved our understanding of the region, but no single model has been accepted yet. This study takes advantage of the dense station networks deployed from France in the north to Canary Islands and Morocco in the south to provide a new high-resolution P-wave velocity model of the structure of the upper-mantle and top of the lower mantle. These images show subvertical small-scale upwellings below Atlas Range, Canary Islands and Central Iberia that seem to cross the transition zone. The results, together with geochemical evidence and a comparison with previous global tomographic models, reveal the ponding or flow of deep-plume material beneath the transition zone, which seems to feed upper-mantle "secondary" pulses. In the upper mantle the plumes, in conjunction with the subduction-related upwellings, allow the hot mantle to rise in the surrounding zones. During its rising, the mantle interacts with horizontal SW slab-driven flow which skirts the Alboran slab and connects with the mantle upwelling below Massif Central through the Valencia Trough rift.

Do deep convolutional neural networks really need to be deep when applied for remote scene classification?

NASA Astrophysics Data System (ADS)

Luo, Chang; Wang, Jie; Feng, Gang; Xu, Suhui; Wang, Shiqiang

2017-10-01

Deep convolutional neural networks (CNNs) have been widely used to obtain high-level representation in various computer vision tasks. However, for remote scene classification, there are not sufficient images to train a very deep CNN from scratch. From two viewpoints of generalization power, we propose two promising kinds of deep CNNs for remote scenes and try to find whether deep CNNs need to be deep for remote scene classification. First, we transfer successful pretrained deep CNNs to remote scenes based on the theory that depth of CNNs brings the generalization power by learning available hypothesis for finite data samples. Second, according to the opposite viewpoint that generalization power of deep CNNs comes from massive memorization and shallow CNNs with enough neural nodes have perfect finite sample expressivity, we design a lightweight deep CNN (LDCNN) for remote scene classification. With five well-known pretrained deep CNNs, experimental results on two independent remote-sensing datasets demonstrate that transferred deep CNNs can achieve state-of-the-art results in an unsupervised setting. However, because of its shallow architecture, LDCNN cannot obtain satisfactory performance, regardless of whether in an unsupervised, semisupervised, or supervised setting. CNNs really need depth to obtain general features for remote scenes. This paper also provides baseline for applying deep CNNs to other remote sensing tasks.

How Deep Neural Networks Can Improve Emotion Recognition on Video Data

DTIC Science & Technology

2016-09-25

HOW DEEP NEURAL NETWORKS CAN IMPROVE EMOTION RECOGNITION ON VIDEO DATA Pooya Khorrami1 , Tom Le Paine1, Kevin Brady2, Charlie Dagli2, Thomas S...this work, we present a system that per- forms emotion recognition on video data using both con- volutional neural networks (CNNs) and recurrent...neural net- works (RNNs). We present our findings on videos from the Audio/Visual+Emotion Challenge (AV+EC2015). In our experiments, we analyze the effects

Single-shot T2 mapping using overlapping-echo detachment planar imaging and a deep convolutional neural network.

PubMed

Cai, Congbo; Wang, Chao; Zeng, Yiqing; Cai, Shuhui; Liang, Dong; Wu, Yawen; Chen, Zhong; Ding, Xinghao; Zhong, Jianhui

2018-04-24

An end-to-end deep convolutional neural network (CNN) based on deep residual network (ResNet) was proposed to efficiently reconstruct reliable T 2 mapping from single-shot overlapping-echo detachment (OLED) planar imaging. The training dataset was obtained from simulations that were carried out on SPROM (Simulation with PRoduct Operator Matrix) software developed by our group. The relationship between the original OLED image containing two echo signals and the corresponding T 2 mapping was learned by ResNet training. After the ResNet was trained, it was applied to reconstruct the T 2 mapping from simulation and in vivo human brain data. Although the ResNet was trained entirely on simulated data, the trained network was generalized well to real human brain data. The results from simulation and in vivo human brain experiments show that the proposed method significantly outperforms the echo-detachment-based method. Reliable T 2 mapping with higher accuracy is achieved within 30 ms after the network has been trained, while the echo-detachment-based OLED reconstruction method took approximately 2 min. The proposed method will facilitate real-time dynamic and quantitative MR imaging via OLED sequence, and deep convolutional neural network has the potential to reconstruct maps from complex MRI sequences efficiently. © 2018 International Society for Magnetic Resonance in Medicine.

Exploring the effects of dimensionality reduction in deep networks for force estimation in robotic-assisted surgery

NASA Astrophysics Data System (ADS)

Aviles, Angelica I.; Alsaleh, Samar; Sobrevilla, Pilar; Casals, Alicia

2016-03-01

Robotic-Assisted Surgery approach overcomes the limitations of the traditional laparoscopic and open surgeries. However, one of its major limitations is the lack of force feedback. Since there is no direct interaction between the surgeon and the tissue, there is no way of knowing how much force the surgeon is applying which can result in irreversible injuries. The use of force sensors is not practical since they impose different constraints. Thus, we make use of a neuro-visual approach to estimate the applied forces, in which the 3D shape recovery together with the geometry of motion are used as input to a deep network based on LSTM-RNN architecture. When deep networks are used in real time, pre-processing of data is a key factor to reduce complexity and improve the network performance. A common pre-processing step is dimensionality reduction which attempts to eliminate redundant and insignificant information by selecting a subset of relevant features to use in model construction. In this work, we show the effects of dimensionality reduction in a real-time application: estimating the applied force in Robotic-Assisted Surgeries. According to the results, we demonstrated positive effects of doing dimensionality reduction on deep networks including: faster training, improved network performance, and overfitting prevention. We also show a significant accuracy improvement, ranging from about 33% to 86%, over existing approaches related to force estimation.

Deforestation and Industrial Forest Patterns in Colombia: a Case Study

NASA Astrophysics Data System (ADS)

Huo, L. Z.; Boschetti, L.; Sparks, A. M.; Clerici, N.

2017-12-01

The recent peace agreement between the government and the Revolutionary Armed Forces of Colombia (FARC) offers new opportunities for peaceful and sustainable development, but at the same time requires a timely effort to protect biological resources, and ecosystem services (Clerici et al., 2016). In this context, we use the 2001-2017 Landsat data record to prototype a methodology to establish a baseline of deforestation, afforestation and industrial forest practices (i.e. establishment and harvest of forest plantations), and to monitor future changes. Two study areas, which have seen considerable deforestation in recent years, were selected: one in the South of the country, at the edge of the Amazon Forest (WRS path 008 row 059) and one in the center, in mixed forest (WRS path 008 row 055). The time series of all the available cloud free Landsat 5, Landsat 7 and Landsat 8 data was classified into a sequence of binary forest/non forest maps using a deep learning model, successfully used in the natural language processing field, trained to detect forest transitions. Recurrent Neural Networks (RNN) is a class of artificial neural network that extends the conventional neural network with loops in the connections (Graves et al., 2013). Unlike a feed-forward neural network, an RNN is able to process the sequential inputs by having a recurrent hidden state whose activation at each step depends on that of the previous steps. In this manner, the RNN provides a good framework to dynamically model time series data, and has been successfully applied to natural language processing in Google (Sutskever et al., 2014). The sequence of forest cover state maps was subsequently post-processed to differentiate between deforestation (e.g. transition from forest to non forest land use) and industrial forest harvest (i.e. timber harvest followed by regrowth), by integrating the detection of temporal patterns, and spatial patterns. References Clerici, N., et al., (2016). Colombia: Dealing in conservation. Science, 354(6309), 190-190. Sutskever I.,et al. (2014). Sequence to sequence learning with neural networks. Advances in neural information processing systems, 3104-3112. Graves A., et al. (2013). Speech recognition with deep recurrent neural networks. In Proc. IEEE Int. Conf. Acoust. Speech Signal Process., 6645-6649.

The Deep Impact Network Experiment Operations Center Monitor and Control System

NASA Technical Reports Server (NTRS)

Wang, Shin-Ywan (Cindy); Torgerson, J. Leigh; Schoolcraft, Joshua; Brenman, Yan

2009-01-01

The Interplanetary Overlay Network (ION) software at JPL is an implementation of Delay/Disruption Tolerant Networking (DTN) which has been proposed as an interplanetary protocol to support space communication. The JPL Deep Impact Network (DINET) is a technology development experiment intended to increase the technical readiness of the JPL implemented ION suite. The DINET Experiment Operations Center (EOC) developed by JPL's Protocol Technology Lab (PTL) was critical in accomplishing the experiment. EOC, containing all end nodes of simulated spaces and one administrative node, exercised publish and subscribe functions for payload data among all end nodes to verify the effectiveness of data exchange over ION protocol stacks. A Monitor and Control System was created and installed on the administrative node as a multi-tiered internet-based Web application to support the Deep Impact Network Experiment by allowing monitoring and analysis of the data delivery and statistics from ION. This Monitor and Control System includes the capability of receiving protocol status messages, classifying and storing status messages into a database from the ION simulation network, and providing web interfaces for viewing the live results in addition to interactive database queries.

On Deep Learning for Trust-Aware Recommendations in Social Networks.

PubMed

Deng, Shuiguang; Huang, Longtao; Xu, Guandong; Wu, Xindong; Wu, Zhaohui

2017-05-01

With the emergence of online social networks, the social network-based recommendation approach is popularly used. The major benefit of this approach is the ability of dealing with the problems with cold-start users. In addition to social networks, user trust information also plays an important role to obtain reliable recommendations. Although matrix factorization (MF) becomes dominant in recommender systems, the recommendation largely relies on the initialization of the user and item latent feature vectors. Aiming at addressing these challenges, we develop a novel trust-based approach for recommendation in social networks. In particular, we attempt to leverage deep learning to determinate the initialization in MF for trust-aware social recommendations and to differentiate the community effect in user's trusted friendships. A two-phase recommendation process is proposed to utilize deep learning in initialization and to synthesize the users' interests and their trusted friends' interests together with the impact of community effect for recommendations. We perform extensive experiments on real-world social network data to demonstrate the accuracy and effectiveness of our proposed approach in comparison with other state-of-the-art methods.

Automatic detection of hemorrhagic pericardial effusion on PMCT using deep learning - a feasibility study.

PubMed

Ebert, Lars C; Heimer, Jakob; Schweitzer, Wolf; Sieberth, Till; Leipner, Anja; Thali, Michael; Ampanozi, Garyfalia

2017-12-01

Post mortem computed tomography (PMCT) can be used as a triage tool to better identify cases with a possibly non-natural cause of death, especially when high caseloads make it impossible to perform autopsies on all cases. Substantial data can be generated by modern medical scanners, especially in a forensic setting where the entire body is documented at high resolution. A solution for the resulting issues could be the use of deep learning techniques for automatic analysis of radiological images. In this article, we wanted to test the feasibility of such methods for forensic imaging by hypothesizing that deep learning methods can detect and segment a hemopericardium in PMCT. For deep learning image analysis software, we used the ViDi Suite 2.0. We retrospectively selected 28 cases with, and 24 cases without, hemopericardium. Based on these data, we trained two separate deep learning networks. The first one classified images into hemopericardium/not hemopericardium, and the second one segmented the blood content. We randomly selected 50% of the data for training and 50% for validation. This process was repeated 20 times. The best performing classification network classified all cases of hemopericardium from the validation images correctly with only a few false positives. The best performing segmentation network would tend to underestimate the amount of blood in the pericardium, which is the case for most networks. This is the first study that shows that deep learning has potential for automated image analysis of radiological images in forensic medicine.

TRMM precipitation analysis of extreme storms in South America: Bias and climatological contribution

NASA Astrophysics Data System (ADS)

Rasmussen, K. L.; Houze, R.; Zuluaga, M. D.; Choi, S. L.; Chaplin, M.

2013-12-01

The TRMM (Tropical Rainfall Measuring Mission) satellite was designed both to measure spatial and temporal variation of tropical rainfall around the globe and to understand the factors controlling the precipitation. TRMM observations have led to the realization that storms just east of the Andes in southeastern South America are among the most intense deep convection in the world. For a complete perspective of the impact of intense precipitation systems on the hydrologic cycle in South America, it is necessary to assess the contribution from various forms of extreme storms to the climatological rainfall. However, recent studies have suggested that the TRMM Precipitation Radar (PR) algorithm significantly underestimates surface rainfall in deep convection over land. Prior to investigating the climatological behavior, this research first investigates the range of the rain bias in storms containing four different types of extreme radar echoes: deep convective cores, deep and wide convective cores, wide convective cores, and broad stratiform regions over South America. The TRMM PR algorithm exhibits bias in all four extreme echo types considered here when the algorithm rates are compared to a range of conventional Z-R relations. Storms with deep convective cores, defined as high reflectivity echo volumes that extend above 10 km in altitude, show the greatest underestimation, and the bias is unrelated to their echo top height. The bias in wide convective cores, defined as high reflectivity echo volumes that extend horizontally over 1,000 km2, relates to the echo top, indicating that storms with significant mixed phase and ice hydrometeors are similarly affected by assumptions in the TRMM PR algorithm. The subtropical region tends to have more intense precipitating systems than the tropics, but the relationship between the TRMM PR rain bias and storm type is the same regardless of the climatological regime. The most extreme storms are typically not collocated with regions of high climatological precipitation. A quantitative approach that accounts for the previously described bias using TRMM PR data is employed to investigate the role of the most extreme precipitating systems on the hydrological cycle in South America. These data are first used to investigate the relative contribution of precipitation from the TRMM-identified echo cores to each separate storm in which the convective cores are embedded. The second part of the study assesses how much of the climatological rainfall in South America is accounted for by storms containing deep convective, wide convective, and broad stratiform echo components. Systems containing these echoes produce very different hydrologic responses. From a hydrologic and climatological viewpoint, this empirical knowledge is critical, as the type of runoff and flooding that may occur depends on the specific character of the convective storm and has broad implications for the hydrological cycle in this region.

Deep thermal structure of Southeast Asia constrained by S-velocity data

NASA Astrophysics Data System (ADS)

Yu, Chuanhai; Shi, Xiaobin; Yang, Xiaoqiu; Zhao, Junfeng; Chen, Mei; Tang, Qunshu

2017-12-01

Southeast Asia, located in the southeastern part of the Eurasian Plate, is surrounded by tectonically active margins, exhibiting intense seismicity and volcanism, contains complex geological units with a perplexing evolution history. Because tectonic evolution is closely related to the deep thermal structure, an accurate estimation of the lithosphere thermal structure and thickness is important in extracting information on tectonics and geodynamics. However, there are significant uncertainties in the calculated deep thermal state constrained only by the observed surface heat flow. In this study, in order to obtain a better-constrained deep thermal state, we first calculate the deep thermal structure of Southeast Asia by employing an empirical relation between S-velocity and temperature, and then we estimate the base of the thermal lithosphere from the calculated temperature-depth profiles. The results show that, in general, the temperature is higher than the dry mantle solidus below the top of the seismic low-velocity zone, possibly indicating the presence of partial melt in the asthenosphere, particularly beneath oceanic basins such as the South China Sea. The temperature at a depth of 80 km in rifted and oceanic basins such as the Thailand Rift Basin, Thailand Bay, Andaman Sea, and South China Sea is about 200 °C higher than in plateaus and subduction zones such as the Khorat Plateau, Sumatra Island, and Philippine Trench regions. We suggest that the relatively cold and thick lithosphere block of the Khorat Plateau has not experienced significant internal deformation and might be extruded and rotated as a rigid block in response to the Indo-Eurasia collision. Our results show that the surface heat flow in the South China Sea is mainly dominated by the deep thermal state. There is a thermal anomaly in the Leiqiong area and in the areas adjacent to the northern margin of the South China Sea, indicating the presence of a high-temperature and thin lithosphere in the area of the well-known and controversial Hainan plume. The thermal lithosphere-asthenosphere boundary uplift area along the Xisha and southeastern Vietnam margin, in the western margin of South China Sea, which corresponds to the volcanic belt around this area, might indicate upwelling of hot mantle materials. The temperature values at 100 and 120 km depths through most regions of Southeast Asia are about 1400-1500 and 1550-1600 °C, respectively, which are nearly uniform with a small temperature difference. Our results also show that the lithosphere becomes thinner from the continent blocks toward the oceanic basins, with the smaller thickness values of 65-70 km in the South China Sea. The estimated base of the lithosphere corresponds approximately to the 1400 °C isotherm and shows good correlation with the tectonic setting.

Deep-water chemosynthetic ecosystem research during the census of marine life decade and beyond: a proposed deep-ocean road map.

PubMed

German, Christopher R; Ramirez-Llodra, Eva; Baker, Maria C; Tyler, Paul A

2011-01-01

The ChEss project of the Census of Marine Life (2002-2010) helped foster internationally-coordinated studies worldwide focusing on exploration for, and characterization of new deep-sea chemosynthetic ecosystem sites. This work has advanced our understanding of the nature and factors controlling the biogeography and biodiversity of these ecosystems in four geographic locations: the Atlantic Equatorial Belt (AEB), the New Zealand region, the Arctic and Antarctic and the SE Pacific off Chile. In the AEB, major discoveries include hydrothermal seeps on the Costa Rica margin, deepest vents found on the Mid-Cayman Rise and the hottest vents found on the Southern Mid-Atlantic Ridge. It was also shown that the major fracture zones on the MAR do not create barriers for the dispersal but may act as trans-Atlantic conduits for larvae. In New Zealand, investigations of a newly found large cold-seep area suggest that this region may be a new biogeographic province. In the Arctic, the newly discovered sites on the Mohns Ridge (71 °N) showed extensive mats of sulfur-oxidisng bacteria, but only one gastropod potentially bears chemosynthetic symbionts, while cold seeps on the Haakon Mossby Mud Volcano (72 °N) are dominated by siboglinid worms. In the Antarctic region, the first hydrothermal vents south of the Polar Front were located and biological results indicate that they may represent a new biogeographic province. The recent exploration of the South Pacific region has provided evidence for a sediment hosted hydrothermal source near a methane-rich cold-seep area. Based on our 8 years of investigations of deep-water chemosynthetic ecosystems worldwide, we suggest highest priorities for future research: (i) continued exploration of the deep-ocean ridge-crest; (ii) increased focus on anthropogenic impacts; (iii) concerted effort to coordinate a major investigation of the deep South Pacific Ocean - the largest contiguous habitat for life within Earth's biosphere, but also the world's least investigated deep-ocean basin.

Deep-Water Chemosynthetic Ecosystem Research during the Census of Marine Life Decade and Beyond: A Proposed Deep-Ocean Road Map

PubMed Central

German, Christopher R.; Ramirez-Llodra, Eva; Baker, Maria C.; Tyler, Paul A.

2011-01-01

The ChEss project of the Census of Marine Life (2002–2010) helped foster internationally-coordinated studies worldwide focusing on exploration for, and characterization of new deep-sea chemosynthetic ecosystem sites. This work has advanced our understanding of the nature and factors controlling the biogeography and biodiversity of these ecosystems in four geographic locations: the Atlantic Equatorial Belt (AEB), the New Zealand region, the Arctic and Antarctic and the SE Pacific off Chile. In the AEB, major discoveries include hydrothermal seeps on the Costa Rica margin, deepest vents found on the Mid-Cayman Rise and the hottest vents found on the Southern Mid-Atlantic Ridge. It was also shown that the major fracture zones on the MAR do not create barriers for the dispersal but may act as trans-Atlantic conduits for larvae. In New Zealand, investigations of a newly found large cold-seep area suggest that this region may be a new biogeographic province. In the Arctic, the newly discovered sites on the Mohns Ridge (71°N) showed extensive mats of sulfur-oxidisng bacteria, but only one gastropod potentially bears chemosynthetic symbionts, while cold seeps on the Haakon Mossby Mud Volcano (72°N) are dominated by siboglinid worms. In the Antarctic region, the first hydrothermal vents south of the Polar Front were located and biological results indicate that they may represent a new biogeographic province. The recent exploration of the South Pacific region has provided evidence for a sediment hosted hydrothermal source near a methane-rich cold-seep area. Based on our 8 years of investigations of deep-water chemosynthetic ecosystems worldwide, we suggest highest priorities for future research: (i) continued exploration of the deep-ocean ridge-crest; (ii) increased focus on anthropogenic impacts; (iii) concerted effort to coordinate a major investigation of the deep South Pacific Ocean – the largest contiguous habitat for life within Earth's biosphere, but also the world's least investigated deep-ocean basin. PMID:21829722

Nematoda from the terrestrial deep subsurface of South Africa.

PubMed

Borgonie, G; García-Moyano, A; Litthauer, D; Bert, W; Bester, A; van Heerden, E; Möller, C; Erasmus, M; Onstott, T C

2011-06-02

Since its discovery over two decades ago, the deep subsurface biosphere has been considered to be the realm of single-cell organisms, extending over three kilometres into the Earth's crust and comprising a significant fraction of the global biosphere. The constraints of temperature, energy, dioxygen and space seemed to preclude the possibility of more-complex, multicellular organisms from surviving at these depths. Here we report species of the phylum Nematoda that have been detected in or recovered from 0.9-3.6-kilometre-deep fracture water in the deep mines of South Africa but have not been detected in the mining water. These subsurface nematodes, including a new species, Halicephalobus mephisto, tolerate high temperature, reproduce asexually and preferentially feed upon subsurface bacteria. Carbon-14 data indicate that the fracture water in which the nematodes reside is 3,000-12,000-year-old palaeometeoric water. Our data suggest that nematodes should be found in other deep hypoxic settings where temperature permits, and that they may control the microbial population density by grazing on fracture surface biofilm patches. Our results expand the known metazoan biosphere and demonstrate that deep ecosystems are more complex than previously accepted. The discovery of multicellular life in the deep subsurface of the Earth also has important implications for the search for subsurface life on other planets in our Solar System.

The design and performance of IceCube DeepCore

NASA Astrophysics Data System (ADS)

Abbasi, R.; Abdou, Y.; Abu-Zayyad, T.; Ackermann, M.; Adams, J.; Aguilar, J. A.; Ahlers, M.; Allen, M. M.; Altmann, D.; Andeen, K.; Auffenberg, J.; Bai, X.; Baker, M.; Barwick, S. W.; Bay, R.; Bazo Alba, J. L.; Beattie, K.; Beatty, J. J.; Bechet, S.; Becker, J. K.; Becker, K.-H.; Benabderrahmane, M. L.; BenZvi, S.; Berdermann, J.; Berghaus, P.; Berley, D.; Bernardini, E.; Bertrand, D.; Besson, D. Z.; Bindig, D.; Bissok, M.; Blaufuss, E.; Blumenthal, J.; Boersma, D. J.; Bohm, C.; Bose, D.; Böser, S.; Botner, O.; Brown, A. M.; Buitink, S.; Caballero-Mora, K. S.; Carson, M.; Chirkin, D.; Christy, B.; Clevermann, F.; Cohen, S.; Colnard, C.; Cowen, D. F.; Cruz Silva, A. H.; D'Agostino, M. V.; Danninger, M.; Daughhetee, J.; Davis, J. C.; De Clercq, C.; Degner, T.; Demirörs, L.; Descamps, F.; Desiati, P.; de Vries-Uiterweerd, G.; DeYoung, T.; Díaz-Vélez, J. C.; Dierckxsens, M.; Dreyer, J.; Dumm, J. P.; Dunkman, M.; Eisch, J.; Ellsworth, R. W.; Engdegård, O.; Euler, S.; Evenson, P. A.; Fadiran, O.; Fazely, A. R.; Fedynitch, A.; Feintzeig, J.; Feusels, T.; Filimonov, K.; Finley, C.; Fischer-Wasels, T.; Fox, B. D.; Franckowiak, A.; Franke, R.; Gaisser, T. K.; Gallagher, J.; Gerhardt, L.; Gladstone, L.; Glüsenkamp, T.; Goldschmidt, A.; Goodman, J. A.; Góra, D.; Grant, D.; Griesel, T.; Groß, A.; Grullon, S.; Gurtner, M.; Ha, C.; Haj Ismail, A.; Hallgren, A.; Halzen, F.; Han, K.; Hanson, K.; Heinen, D.; Helbing, K.; Hellauer, R.; Hickford, S.; Hill, G. C.; Hoffman, K. D.; Hoffmann, B.; Homeier, A.; Hoshina, K.; Huelsnitz, W.; Hülß, J.-P.; Hulth, P. O.; Hultqvist, K.; Hussain, S.; Ishihara, A.; Jacobi, E.; Jacobsen, J.; Japaridze, G. S.; Johansson, H.; Kampert, K.-H.; Kappes, A.; Karg, T.; Karle, A.; Kenny, P.; Kiryluk, J.; Kislat, F.; Klein, S. R.; Köhne, J.-H.; Kohnen, G.; Kolanoski, H.; Köpke, L.; Koskinen, D. J.; Kowalski, M.; Kowarik, T.; Krasberg, M.; Kroll, G.; Kurahashi, N.; Kuwabara, T.; Labare, M.; Laihem, K.; Landsman, H.; Larson, M. J.; Lauer, R.; Lünemann, J.; Madsen, J.; Marotta, A.; Maruyama, R.; Mase, K.; Matis, H. S.; Meagher, K.; Merck, M.; Mészáros, P.; Meures, T.; Miarecki, S.; Middell, E.; Milke, N.; Miller, J.; Montaruli, T.; Morse, R.; Movit, S. M.; Nahnhauer, R.; Nam, J. W.; Naumann, U.; Nygren, D. R.; Odrowski, S.; Olivas, A.; Olivo, M.; O'Murchadha, A.; Panknin, S.; Paul, L.; Pérez de los Heros, C.; Petrovic, J.; Piegsa, A.; Pieloth, D.; Porrata, R.; Posselt, J.; Price, P. B.; Przybylski, G. T.; Rawlins, K.; Redl, P.; Resconi, E.; Rhode, W.; Ribordy, M.; Richman, M.; Rodrigues, J. P.; Rothmaier, F.; Rott, C.; Ruhe, T.; Rutledge, D.; Ruzybayev, B.; Ryckbosch, D.; Sander, H.-G.; Santander, M.; Sarkar, S.; Schatto, K.; Schmidt, T.; Schönwald, A.; Schukraft, A.; Schultes, A.; Schulz, O.; Schunck, M.; Seckel, D.; Semburg, B.; Seo, S. H.; Sestayo, Y.; Seunarine, S.; Silvestri, A.; Spiczak, G. M.; Spiering, C.; Stamatikos, M.; Stanev, T.; Stezelberger, T.; Stokstad, R. G.; Stößl, A.; Strahler, E. A.; Ström, R.; Stüer, M.; Sullivan, G. W.; Swillens, Q.; Taavola, H.; Taboada, I.; Tamburro, A.; Tepe, A.; Ter-Antonyan, S.; Tilav, S.; Toale, P. A.; Toscano, S.; Tosi, D.; van Eijndhoven, N.; Vandenbroucke, J.; Van Overloop, A.; van Santen, J.; Vehring, M.; Voge, M.; Walck, C.; Waldenmaier, T.; Wallraff, M.; Walter, M.; Weaver, Ch.; Wendt, C.; Westerhoff, S.; Whitehorn, N.; Wiebe, K.; Wiebusch, C. H.; Williams, D. R.; Wischnewski, R.; Wissing, H.; Wolf, M.; Wood, T. R.; Woschnagg, K.; Xu, C.; Xu, D. L.; Xu, X. W.; Yanez, J. P.; Yodh, G.; Yoshida, S.; Zarzhitsky, P.; Zoll, M.

2012-05-01

The IceCube neutrino observatory in operation at the South Pole, Antarctica, comprises three distinct components: a large buried array for ultrahigh energy neutrino detection, a surface air shower array, and a new buried component called DeepCore. DeepCore was designed to lower the IceCube neutrino energy threshold by over an order of magnitude, to energies as low as about 10 GeV. DeepCore is situated primarily 2100 m below the surface of the icecap at the South Pole, at the bottom center of the existing IceCube array, and began taking physics data in May 2010. Its location takes advantage of the exceptionally clear ice at those depths and allows it to use the surrounding IceCube detector as a highly efficient active veto against the principal background of downward-going muons produced in cosmic-ray air showers. DeepCore has a module density roughly five times higher than that of the standard IceCube array, and uses photomultiplier tubes with a new photocathode featuring a quantum efficiency about 35% higher than standard IceCube PMTs. Taken together, these features of DeepCore will increase IceCube's sensitivity to neutrinos from WIMP dark matter annihilations, atmospheric neutrino oscillations, galactic supernova neutrinos, and point sources of neutrinos in the northern and southern skies. In this paper we describe the design and initial performance of DeepCore.

The Design and Performance of IceCube DeepCore

NASA Technical Reports Server (NTRS)

Stamatikos, M.

2012-01-01

The IceCube neutrino observatory in operation at the South Pole, Antarctica, comprises three distinct components: a large buried array for ultrahigh energy neutrino detection, a surface air shower array, and a new buried component called DeepCore. DeepCore was designed to lower the IceCube neutrino energy threshold by over an order of magnitude, to energies as low as about 10 GeV. DeepCore is situated primarily 2100 m below the surface of the icecap at the South Pole, at the bottom center of the existing IceCube array, and began taking pbysics data in May 2010. Its location takes advantage of the exceptionally clear ice at those depths and allows it to use the surrounding IceCube detector as a highly efficient active veto against the principal background of downward-going muons produced in cosmic-ray air showers. DeepCore has a module density roughly five times higher than that of the standard IceCube array, and uses photomultiplier tubes with a new photocathode featuring a quantum efficiency about 35% higher than standard IceCube PMTs. Taken together, these features of DeepCore will increase IceCube's sensitivity to neutrinos from WIMP dark matter annihilations, atmospheric neutrino oscillations, galactic supernova neutrinos, and point sources of neutrinos in the northern and southern skies. In this paper we describe the design and initial performance of DeepCore.

Nonlinear Deep Kernel Learning for Image Annotation.

PubMed

Jiu, Mingyuan; Sahbi, Hichem

2017-02-08

Multiple kernel learning (MKL) is a widely used technique for kernel design. Its principle consists in learning, for a given support vector classifier, the most suitable convex (or sparse) linear combination of standard elementary kernels. However, these combinations are shallow and often powerless to capture the actual similarity between highly semantic data, especially for challenging classification tasks such as image annotation. In this paper, we redefine multiple kernels using deep multi-layer networks. In this new contribution, a deep multiple kernel is recursively defined as a multi-layered combination of nonlinear activation functions, each one involves a combination of several elementary or intermediate kernels, and results into a positive semi-definite deep kernel. We propose four different frameworks in order to learn the weights of these networks: supervised, unsupervised, kernel-based semisupervised and Laplacian-based semi-supervised. When plugged into support vector machines (SVMs), the resulting deep kernel networks show clear gain, compared to several shallow kernels for the task of image annotation. Extensive experiments and analysis on the challenging ImageCLEF photo annotation benchmark, the COREL5k database and the Banana dataset validate the effectiveness of the proposed method.

Deep learning with convolutional neural network in radiology.

PubMed

Yasaka, Koichiro; Akai, Hiroyuki; Kunimatsu, Akira; Kiryu, Shigeru; Abe, Osamu

2018-04-01

Deep learning with a convolutional neural network (CNN) is gaining attention recently for its high performance in image recognition. Images themselves can be utilized in a learning process with this technique, and feature extraction in advance of the learning process is not required. Important features can be automatically learned. Thanks to the development of hardware and software in addition to techniques regarding deep learning, application of this technique to radiological images for predicting clinically useful information, such as the detection and the evaluation of lesions, etc., are beginning to be investigated. This article illustrates basic technical knowledge regarding deep learning with CNNs along the actual course (collecting data, implementing CNNs, and training and testing phases). Pitfalls regarding this technique and how to manage them are also illustrated. We also described some advanced topics of deep learning, results of recent clinical studies, and the future directions of clinical application of deep learning techniques.

Oblique view to south OvertheHorizon Backscatter Radar Network, Mountain ...

Library of Congress Historic Buildings Survey, Historic Engineering Record, Historic Landscapes Survey

Oblique view to south - Over-the-Horizon Backscatter Radar Network, Mountain Home Air Force Operations Building, On Desert Street at 9th Avenue Mountain Home Air Force Base, Mountain Home, Elmore County, ID

Classification of breast MRI lesions using small-size training sets: comparison of deep learning approaches

NASA Astrophysics Data System (ADS)

Amit, Guy; Ben-Ari, Rami; Hadad, Omer; Monovich, Einat; Granot, Noa; Hashoul, Sharbell

2017-03-01

Diagnostic interpretation of breast MRI studies requires meticulous work and a high level of expertise. Computerized algorithms can assist radiologists by automatically characterizing the detected lesions. Deep learning approaches have shown promising results in natural image classification, but their applicability to medical imaging is limited by the shortage of large annotated training sets. In this work, we address automatic classification of breast MRI lesions using two different deep learning approaches. We propose a novel image representation for dynamic contrast enhanced (DCE) breast MRI lesions, which combines the morphological and kinetics information in a single multi-channel image. We compare two classification approaches for discriminating between benign and malignant lesions: training a designated convolutional neural network and using a pre-trained deep network to extract features for a shallow classifier. The domain-specific trained network provided higher classification accuracy, compared to the pre-trained model, with an area under the ROC curve of 0.91 versus 0.81, and an accuracy of 0.83 versus 0.71. Similar accuracy was achieved in classifying benign lesions, malignant lesions, and normal tissue images. The trained network was able to improve accuracy by using the multi-channel image representation, and was more robust to reductions in the size of the training set. A small-size convolutional neural network can learn to accurately classify findings in medical images using only a few hundred images from a few dozen patients. With sufficient data augmentation, such a network can be trained to outperform a pre-trained out-of-domain classifier. Developing domain-specific deep-learning models for medical imaging can facilitate technological advancements in computer-aided diagnosis.

Cough event classification by pretrained deep neural network.

PubMed

Liu, Jia-Ming; You, Mingyu; Wang, Zheng; Li, Guo-Zheng; Xu, Xianghuai; Qiu, Zhongmin

2015-01-01

Cough is an essential symptom in respiratory diseases. In the measurement of cough severity, an accurate and objective cough monitor is expected by respiratory disease society. This paper aims to introduce a better performed algorithm, pretrained deep neural network (DNN), to the cough classification problem, which is a key step in the cough monitor. The deep neural network models are built from two steps, pretrain and fine-tuning, followed by a Hidden Markov Model (HMM) decoder to capture tamporal information of the audio signals. By unsupervised pretraining a deep belief network, a good initialization for a deep neural network is learned. Then the fine-tuning step is a back propogation tuning the neural network so that it can predict the observation probability associated with each HMM states, where the HMM states are originally achieved by force-alignment with a Gaussian Mixture Model Hidden Markov Model (GMM-HMM) on the training samples. Three cough HMMs and one noncough HMM are employed to model coughs and noncoughs respectively. The final decision is made based on viterbi decoding algorihtm that generates the most likely HMM sequence for each sample. A sample is labeled as cough if a cough HMM is found in the sequence. The experiments were conducted on a dataset that was collected from 22 patients with respiratory diseases. Patient dependent (PD) and patient independent (PI) experimental settings were used to evaluate the models. Five criteria, sensitivity, specificity, F1, macro average and micro average are shown to depict different aspects of the models. From overall evaluation criteria, the DNN based methods are superior to traditional GMM-HMM based method on F1 and micro average with maximal 14% and 11% error reduction in PD and 7% and 10% in PI, meanwhile keep similar performances on macro average. They also surpass GMM-HMM model on specificity with maximal 14% error reduction on both PD and PI. In this paper, we tried pretrained deep neural network in cough classification problem. Our results showed that comparing with the conventional GMM-HMM framework, the HMM-DNN could get better overall performance on cough classification task.

Deep learning in color: towards automated quark/gluon jet discrimination

DOE PAGES

Komiske, Patrick T.; Metodiev, Eric M.; Schwartz, Matthew D.

2017-01-25

Artificial intelligence offers the potential to automate challenging data-processing tasks in collider physics. Here, to establish its prospects, we explore to what extent deep learning with convolutional neural networks can discriminate quark and gluon jets better than observables designed by physicists. Our approach builds upon the paradigm that a jet can be treated as an image, with intensity given by the local calorimeter deposits. We supplement this construction by adding color to the images, with red, green and blue intensities given by the transverse momentum in charged particles, transverse momentum in neutral particles, and pixel-level charged particle counts. Overall, themore » deep networks match or outperform traditional jet variables. We also find that, while various simulations produce different quark and gluon jets, the neural networks are surprisingly insensitive to these differences, similar to traditional observables. This suggests that the networks can extract robust physical information from imperfect simulations.« less

Deep learning in color: towards automated quark/gluon jet discrimination

NASA Astrophysics Data System (ADS)

Komiske, Patrick T.; Metodiev, Eric M.; Schwartz, Matthew D.

2017-01-01

Artificial intelligence offers the potential to automate challenging data-processing tasks in collider physics. To establish its prospects, we explore to what extent deep learning with convolutional neural networks can discriminate quark and gluon jets better than observables designed by physicists. Our approach builds upon the paradigm that a jet can be treated as an image, with intensity given by the local calorimeter deposits. We supplement this construction by adding color to the images, with red, green and blue intensities given by the transverse momentum in charged particles, transverse momentum in neutral particles, and pixel-level charged particle counts. Overall, the deep networks match or outperform traditional jet variables. We also find that, while various simulations produce different quark and gluon jets, the neural networks are surprisingly insensitive to these differences, similar to traditional observables. This suggests that the networks can extract robust physical information from imperfect simulations.

Deep learning in color: towards automated quark/gluon jet discrimination

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

Komiske, Patrick T.; Metodiev, Eric M.; Schwartz, Matthew D.

Artificial intelligence offers the potential to automate challenging data-processing tasks in collider physics. Here, to establish its prospects, we explore to what extent deep learning with convolutional neural networks can discriminate quark and gluon jets better than observables designed by physicists. Our approach builds upon the paradigm that a jet can be treated as an image, with intensity given by the local calorimeter deposits. We supplement this construction by adding color to the images, with red, green and blue intensities given by the transverse momentum in charged particles, transverse momentum in neutral particles, and pixel-level charged particle counts. Overall, themore » deep networks match or outperform traditional jet variables. We also find that, while various simulations produce different quark and gluon jets, the neural networks are surprisingly insensitive to these differences, similar to traditional observables. This suggests that the networks can extract robust physical information from imperfect simulations.« less

Automated Deep Learning-Based System to Identify Endothelial Cells Derived from Induced Pluripotent Stem Cells.

PubMed

Kusumoto, Dai; Lachmann, Mark; Kunihiro, Takeshi; Yuasa, Shinsuke; Kishino, Yoshikazu; Kimura, Mai; Katsuki, Toshiomi; Itoh, Shogo; Seki, Tomohisa; Fukuda, Keiichi

2018-06-05

Deep learning technology is rapidly advancing and is now used to solve complex problems. Here, we used deep learning in convolutional neural networks to establish an automated method to identify endothelial cells derived from induced pluripotent stem cells (iPSCs), without the need for immunostaining or lineage tracing. Networks were trained to predict whether phase-contrast images contain endothelial cells based on morphology only. Predictions were validated by comparison to immunofluorescence staining for CD31, a marker of endothelial cells. Method parameters were then automatically and iteratively optimized to increase prediction accuracy. We found that prediction accuracy was correlated with network depth and pixel size of images to be analyzed. Finally, K-fold cross-validation confirmed that optimized convolutional neural networks can identify endothelial cells with high performance, based only on morphology. Copyright © 2018 The Author(s). Published by Elsevier Inc. All rights reserved.

The applications of deep neural networks to sdBV classification

NASA Astrophysics Data System (ADS)

Boudreaux, Thomas M.

2017-12-01

With several new large-scale surveys on the horizon, including LSST, TESS, ZTF, and Evryscope, faster and more accurate analysis methods will be required to adequately process the enormous amount of data produced. Deep learning, used in industry for years now, allows for advanced feature detection in minimally prepared datasets at very high speeds; however, despite the advantages of this method, its application to astrophysics has not yet been extensively explored. This dearth may be due to a lack of training data available to researchers. Here we generate synthetic data loosely mimicking the properties of acoustic mode pulsating stars and we show that two separate paradigms of deep learning - the Artificial Neural Network And the Convolutional Neural Network - can both be used to classify this synthetic data effectively. And that additionally this classification can be performed at relatively high levels of accuracy with minimal time spent adjusting network hyperparameters.

Deep learning of orthographic representations in baboons.

PubMed

Hannagan, Thomas; Ziegler, Johannes C; Dufau, Stéphane; Fagot, Joël; Grainger, Jonathan

2014-01-01

What is the origin of our ability to learn orthographic knowledge? We use deep convolutional networks to emulate the primate's ventral visual stream and explore the recent finding that baboons can be trained to discriminate English words from nonwords. The networks were exposed to the exact same sequence of stimuli and reinforcement signals as the baboons in the experiment, and learned to map real visual inputs (pixels) of letter strings onto binary word/nonword responses. We show that the networks' highest levels of representations were indeed sensitive to letter combinations as postulated in our previous research. The model also captured the key empirical findings, such as generalization to novel words, along with some intriguing inter-individual differences. The present work shows the merits of deep learning networks that can simulate the whole processing chain all the way from the visual input to the response while allowing researchers to analyze the complex representations that emerge during the learning process.

Identification of Abnormal System Noise Temperature Patterns in Deep Space Network Antennas Using Neural Network Trained Fuzzy Logic

NASA Technical Reports Server (NTRS)

Lu, Thomas; Pham, Timothy; Liao, Jason

2011-01-01

This paper presents the development of a fuzzy logic function trained by an artificial neural network to classify the system noise temperature (SNT) of antennas in the NASA Deep Space Network (DSN). The SNT data were classified into normal, marginal, and abnormal classes. The irregular SNT pattern was further correlated with link margin and weather data. A reasonably good correlation is detected among high SNT, low link margin and the effect of bad weather; however we also saw some unexpected non-correlations which merit further study in the future.

Temperature control simulation for a microwave transmitter cooling system. [deep space network

NASA Technical Reports Server (NTRS)

Yung, C. S.

1980-01-01

The thermal performance of a temperature control system for the antenna microwave transmitter (klystron tube) of the Deep Space Network antenna tracking system is discussed. In particular the mathematical model is presented along with the details of a computer program which is written for the system simulation and the performance parameterization. Analytical expressions are presented.

Gas Classification Using Deep Convolutional Neural Networks.

PubMed

Peng, Pai; Zhao, Xiaojin; Pan, Xiaofang; Ye, Wenbin

2018-01-08

In this work, we propose a novel Deep Convolutional Neural Network (DCNN) tailored for gas classification. Inspired by the great success of DCNN in the field of computer vision, we designed a DCNN with up to 38 layers. In general, the proposed gas neural network, named GasNet, consists of: six convolutional blocks, each block consist of six layers; a pooling layer; and a fully-connected layer. Together, these various layers make up a powerful deep model for gas classification. Experimental results show that the proposed DCNN method is an effective technique for classifying electronic nose data. We also demonstrate that the DCNN method can provide higher classification accuracy than comparable Support Vector Machine (SVM) methods and Multiple Layer Perceptron (MLP).

Gas Classification Using Deep Convolutional Neural Networks

PubMed Central

Peng, Pai; Zhao, Xiaojin; Pan, Xiaofang; Ye, Wenbin

2018-01-01

In this work, we propose a novel Deep Convolutional Neural Network (DCNN) tailored for gas classification. Inspired by the great success of DCNN in the field of computer vision, we designed a DCNN with up to 38 layers. In general, the proposed gas neural network, named GasNet, consists of: six convolutional blocks, each block consist of six layers; a pooling layer; and a fully-connected layer. Together, these various layers make up a powerful deep model for gas classification. Experimental results show that the proposed DCNN method is an effective technique for classifying electronic nose data. We also demonstrate that the DCNN method can provide higher classification accuracy than comparable Support Vector Machine (SVM) methods and Multiple Layer Perceptron (MLP). PMID:29316723

Implementation of an Antenna Array Signal Processing Breadboard for the Deep Space Network

NASA Technical Reports Server (NTRS)

Navarro, Robert

2006-01-01

The Deep Space Network Large Array will replace/augment 34 and 70 meter antenna assets. The array will mainly be used to support NASA's deep space telemetry, radio science, and navigation requirements. The array project will deploy three complexes in the western U.S., Australia, and European longitude each with 400 12m downlink antennas and a DSN central facility at JPL. THis facility will remotely conduct all real-time monitor and control for the network. Signal processing objectives include: provide a means to evaluate the performance of the Breadboard Array's antenna subsystem; design and build prototype hardware; demonstrate and evaluate proposed signal processing techniques; and gain experience with various technologies that may be used in the Large Array. Results are summarized..

32 GHz Celestial Reference Frame Survey for Dec < -45 deg.

NASA Astrophysics Data System (ADS)

Horiuchi, Shinji; Phillips, Chris; Stevens, Jamie; Jacobs, Christopher; Sotuela, Ioana; Garcia miro, Cristina

2013-04-01

(We resubmit this proposal to extend from the previous semester. The 24 hour blocks for ATCA and Mopra were granted in May 2012 but canceled because fringe test before the scheduled experiment failed although fringes were detected between Mopra and Tidbinbilla. As it turned out ATCA had an issue with frequency standard, which has now been resolved.) We propose to conduct a LBA survey of compact radio sources at 32 GHz near the south pole region. This is the first attempt to fill the gap in the existing 32 GHz catalogue establish by NASA Deep Space Network toward completing the full sky celestial reference frame at 32 GHz. The catalogue will be used for future spacecraft navigation by NASA and other space agencies as well as for radio astronomical observations with southern radio telescope arrays such as ATCA and LBA.

Using Organizational Network Analysis to Plan Cancer Screening Programs for Vulnerable Populations

PubMed Central

Carothers, Bobbi J.; Lofters, Aisha K.

2014-01-01

Objectives. We examined relationships among organizations in a cancer screening network to inform the development of interventions to improve cancer screening for South Asians living in the Peel region of Ontario. Methods. From April to July 2012, we surveyed decision-makers, program managers, and program staff in 22 organizations in the South Asian cancer screening network in the Peel region. We used a network analytic approach to evaluate density (range = 0%–100%, number of ties among organizations in the network expressed as a percentage of all possible ties), centralization (range = 0–1, the extent of variability in centrality), and node characteristics for the communication, collaboration, and referral networks. Results. Density was similar across communication (15%), collaboration (17%), and referral (19%) networks. Centralization was greater in the collaboration network (0.30) than the communication network (0.24), and degree centralization was greater in the inbound (0.42) than the outbound (0.37) referral network. Diverse organizations were central to the networks. Conclusions. Certain organizations were unexpectedly important to the South Asian cancer screening network. Program planning was informed by identifying opportunities to strengthen linkages between key organizations and to leverage existing ties. PMID:24328613

Using organizational network analysis to plan cancer screening programs for vulnerable populations.

PubMed

Lobb, Rebecca; Carothers, Bobbi J; Lofters, Aisha K

2014-02-01

We examined relationships among organizations in a cancer screening network to inform the development of interventions to improve cancer screening for South Asians living in the Peel region of Ontario. From April to July 2012, we surveyed decision-makers, program managers, and program staff in 22 organizations in the South Asian cancer screening network in the Peel region. We used a network analytic approach to evaluate density (range = 0%-100%, number of ties among organizations in the network expressed as a percentage of all possible ties), centralization (range = 0-1, the extent of variability in centrality), and node characteristics for the communication, collaboration, and referral networks. Density was similar across communication (15%), collaboration (17%), and referral (19%) networks. Centralization was greater in the collaboration network (0.30) than the communication network (0.24), and degree centralization was greater in the inbound (0.42) than the outbound (0.37) referral network. Diverse organizations were central to the networks. Certain organizations were unexpectedly important to the South Asian cancer screening network. Program planning was informed by identifying opportunities to strengthen linkages between key organizations and to leverage existing ties.

A climatology of potential severe convective environments across South Africa

NASA Astrophysics Data System (ADS)

Blamey, R. C.; Middleton, C.; Lennard, C.; Reason, C. J. C.

2017-09-01

Severe thunderstorms pose a considerable risk to society and the economy of South Africa during the austral summer months (October-March). Yet, the frequency and distribution of such severe storms is poorly understood, which partly stems out of an inadequate observation network. Given the lack of observations, alternative methods have focused on the relationship between severe storms and their associated environments. One such approach is to use a combination of covariant discriminants, derived from gridded datasets, as a probabilistic proxy for the development of severe storms. These covariates describe some key ingredient for severe convective storm development, such as the presence of instability. Using a combination of convective available potential energy and deep-layer vertical shear from Climate Forecast System Reanalysis, this study establishes a climatology of potential severe convective environments across South Africa for the period 1979-2010. Results indicate that early austral summer months are most likely associated with conditions that are conducive to the development of severe storms over the interior of South Africa. The east coast of the country is a hotspot for potential severe convective environments throughout the summer months. This is likely due to the close proximity of the Agulhas Current, which produces high latent heat fluxes and acts as a key moisture source. No obvious relationship is established between the frequency of potential severe convective environments and the main large-scale modes of variability in the Southern Hemisphere, such as ENSO. This implies that several factors, possibly more localised, may modulate the spatial and temporal frequency of severe thunderstorms across the region.

The effects of deep network topology on mortality prediction.

PubMed

Hao Du; Ghassemi, Mohammad M; Mengling Feng

2016-08-01

Deep learning has achieved remarkable results in the areas of computer vision, speech recognition, natural language processing and most recently, even playing Go. The application of deep-learning to problems in healthcare, however, has gained attention only in recent years, and it's ultimate place at the bedside remains a topic of skeptical discussion. While there is a growing academic interest in the application of Machine Learning (ML) techniques to clinical problems, many in the clinical community see little incentive to upgrade from simpler methods, such as logistic regression, to deep learning. Logistic regression, after all, provides odds ratios, p-values and confidence intervals that allow for ease of interpretation, while deep nets are often seen as `black-boxes' which are difficult to understand and, as of yet, have not demonstrated performance levels far exceeding their simpler counterparts. If deep learning is to ever take a place at the bedside, it will require studies which (1) showcase the performance of deep-learning methods relative to other approaches and (2) interpret the relationships between network structure, model performance, features and outcomes. We have chosen these two requirements as the goal of this study. In our investigation, we utilized a publicly available EMR dataset of over 32,000 intensive care unit patients and trained a Deep Belief Network (DBN) to predict patient mortality at discharge. Utilizing an evolutionary algorithm, we demonstrate automated topology selection for DBNs. We demonstrate that with the correct topology selection, DBNs can achieve better prediction performance compared to several bench-marking methods.

A metagenomic window into carbon metabolism at 3 km depth in Precambrian continental crust

PubMed Central

Magnabosco, Cara; Ryan, Kathleen; Lau, Maggie C Y; Kuloyo, Olukayode; Sherwood Lollar, Barbara; Kieft, Thomas L; van Heerden, Esta; Onstott, Tullis C

2016-01-01

Subsurface microbial communities comprise a significant fraction of the global prokaryotic biomass; however, the carbon metabolisms that support the deep biosphere have been relatively unexplored. In order to determine the predominant carbon metabolisms within a 3-km deep fracture fluid system accessed via the Tau Tona gold mine (Witwatersrand Basin, South Africa), metagenomic and thermodynamic analyses were combined. Within our system of study, the energy-conserving reductive acetyl-CoA (Wood-Ljungdahl) pathway was found to be the most abundant carbon fixation pathway identified in the metagenome. Carbon monoxide dehydrogenase genes that have the potential to participate in (1) both autotrophic and heterotrophic metabolisms through the reversible oxidization of CO and subsequent transfer of electrons for sulfate reduction, (2) direct utilization of H2 and (3) methanogenesis were identified. The most abundant members of the metagenome belonged to Euryarchaeota (22%) and Firmicutes (57%)—by far, the highest relative abundance of Euryarchaeota yet reported from deep fracture fluids in South Africa and one of only five Firmicutes-dominated deep fracture fluids identified in the region. Importantly, by combining the metagenomics data and thermodynamic modeling of this study with previously published isotopic and community composition data from the South African subsurface, we are able to demonstrate that Firmicutes-dominated communities are associated with a particular hydrogeologic environment, specifically the older, more saline and more reducing waters. PMID:26325359

Social Networks of Adults with an Intellectual Disability from South Asian and White Communities in the United Kingdom: A Comparison

ERIC Educational Resources Information Center

Bhardwaj, Anjali K.; Forrester-Jones, Rachel V. E.; Murphy, Glynis H.

2018-01-01

Background: Little research exists comparing the social networks of people with intellectual disability (ID) from South Asian and White backgrounds. This UK study reports on the barriers that South Asian people with intellectual disability face in relation to social inclusion compared to their White counterparts. Materials and methods: A…

A Robust Deep-Learning-Based Detector for Real-Time Tomato Plant Diseases and Pests Recognition.

PubMed

Fuentes, Alvaro; Yoon, Sook; Kim, Sang Cheol; Park, Dong Sun

2017-09-04

Plant Diseases and Pests are a major challenge in the agriculture sector. An accurate and a faster detection of diseases and pests in plants could help to develop an early treatment technique while substantially reducing economic losses. Recent developments in Deep Neural Networks have allowed researchers to drastically improve the accuracy of object detection and recognition systems. In this paper, we present a deep-learning-based approach to detect diseases and pests in tomato plants using images captured in-place by camera devices with various resolutions. Our goal is to find the more suitable deep-learning architecture for our task. Therefore, we consider three main families of detectors: Faster Region-based Convolutional Neural Network (Faster R-CNN), Region-based Fully Convolutional Network (R-FCN), and Single Shot Multibox Detector (SSD), which for the purpose of this work are called "deep learning meta-architectures". We combine each of these meta-architectures with "deep feature extractors" such as VGG net and Residual Network (ResNet). We demonstrate the performance of deep meta-architectures and feature extractors, and additionally propose a method for local and global class annotation and data augmentation to increase the accuracy and reduce the number of false positives during training. We train and test our systems end-to-end on our large Tomato Diseases and Pests Dataset, which contains challenging images with diseases and pests, including several inter- and extra-class variations, such as infection status and location in the plant. Experimental results show that our proposed system can effectively recognize nine different types of diseases and pests, with the ability to deal with complex scenarios from a plant's surrounding area.

Use of an Artificial Neural Network to Construct a Model of Predicting Deep Fungal Infection in Lung Cancer Patients.

PubMed

Chen, Jian; Chen, Jie; Ding, Hong-Yan; Pan, Qin-Shi; Hong, Wan-Dong; Xu, Gang; Yu, Fang-You; Wang, Yu-Min

2015-01-01

The statistical methods to analyze and predict the related dangerous factors of deep fungal infection in lung cancer patients were several, such as logic regression analysis, meta-analysis, multivariate Cox proportional hazards model analysis, retrospective analysis, and so on, but the results are inconsistent. A total of 696 patients with lung cancer were enrolled. The factors were compared employing Student's t-test or the Mann-Whitney test or the Chi-square test and variables that were significantly related to the presence of deep fungal infection selected as candidates for input into the final artificial neural network analysis (ANN) model. The receiver operating characteristic (ROC) and area under curve (AUC) were used to evaluate the performance of the artificial neural network (ANN) model and logistic regression (LR) model. The prevalence of deep fungal infection from lung cancer in this entire study population was 32.04%(223/696), deep fungal infections occur in sputum specimens 44.05% (200/454). The ratio of candida albicans was 86.99% (194/223) in the total fungi. It was demonstrated that older (≥65 years), use of antibiotics, low serum albumin concentrations (≤37.18 g /L), radiotherapy, surgery, low hemoglobin hyperlipidemia (≤93.67 g /L), long time of hospitalization (≥14 days) were apt to deep fungal infection and the ANN model consisted of the seven factors. The AUC of ANN model (0.829±0.019) was higher than that of LR model (0.756±0.021). The artificial neural network model with variables consisting of age, use of antibiotics, serum albumin concentrations, received radiotherapy, received surgery, hemoglobin, time of hospitalization should be useful for predicting the deep fungal infection in lung cancer.

A Robust Deep-Learning-Based Detector for Real-Time Tomato Plant Diseases and Pests Recognition

PubMed Central

Yoon, Sook; Kim, Sang Cheol; Park, Dong Sun

2017-01-01

Plant Diseases and Pests are a major challenge in the agriculture sector. An accurate and a faster detection of diseases and pests in plants could help to develop an early treatment technique while substantially reducing economic losses. Recent developments in Deep Neural Networks have allowed researchers to drastically improve the accuracy of object detection and recognition systems. In this paper, we present a deep-learning-based approach to detect diseases and pests in tomato plants using images captured in-place by camera devices with various resolutions. Our goal is to find the more suitable deep-learning architecture for our task. Therefore, we consider three main families of detectors: Faster Region-based Convolutional Neural Network (Faster R-CNN), Region-based Fully Convolutional Network (R-FCN), and Single Shot Multibox Detector (SSD), which for the purpose of this work are called “deep learning meta-architectures”. We combine each of these meta-architectures with “deep feature extractors” such as VGG net and Residual Network (ResNet). We demonstrate the performance of deep meta-architectures and feature extractors, and additionally propose a method for local and global class annotation and data augmentation to increase the accuracy and reduce the number of false positives during training. We train and test our systems end-to-end on our large Tomato Diseases and Pests Dataset, which contains challenging images with diseases and pests, including several inter- and extra-class variations, such as infection status and location in the plant. Experimental results show that our proposed system can effectively recognize nine different types of diseases and pests, with the ability to deal with complex scenarios from a plant’s surrounding area. PMID:28869539

Generating Seismograms with Deep Neural Networks

NASA Astrophysics Data System (ADS)

Krischer, L.; Fichtner, A.

2017-12-01

The recent surge of successful uses of deep neural networks in computer vision, speech recognition, and natural language processing, mainly enabled by the availability of fast GPUs and extremely large data sets, is starting to see many applications across all natural sciences. In seismology these are largely confined to classification and discrimination tasks. In this contribution we explore the use of deep neural networks for another class of problems: so called generative models.Generative modelling is a branch of statistics concerned with generating new observed data samples, usually by drawing from some underlying probability distribution. Samples with specific attributes can be generated by conditioning on input variables. In this work we condition on seismic source (mechanism and location) and receiver (location) parameters to generate multi-component seismograms.The deep neural networks are trained on synthetic data calculated with Instaseis (http://instaseis.net, van Driel et al. (2015)) and waveforms from the global ShakeMovie project (http://global.shakemovie.princeton.edu, Tromp et al. (2010)). The underlying radially symmetric or smoothly three dimensional Earth structures result in comparatively small waveform differences from similar events or at close receivers and the networks learn to interpolate between training data samples.Of particular importance is the chosen misfit functional. Generative adversarial networks (Goodfellow et al. (2014)) implement a system in which two networks compete: the generator network creates samples and the discriminator network distinguishes these from the true training examples. Both are trained in an adversarial fashion until the discriminator can no longer distinguish between generated and real samples. We show how this can be applied to seismograms and in particular how it compares to networks trained with more conventional misfit metrics. Last but not least we attempt to shed some light on the black-box nature of neural networks by estimating the quality and uncertainties of the generated seismograms.

A Risk Stratification Model for Lung Cancer Based on Gene Coexpression Network and Deep Learning

PubMed Central

2018-01-01

Risk stratification model for lung cancer with gene expression profile is of great interest. Instead of previous models based on individual prognostic genes, we aimed to develop a novel system-level risk stratification model for lung adenocarcinoma based on gene coexpression network. Using multiple microarray, gene coexpression network analysis was performed to identify survival-related networks. A deep learning based risk stratification model was constructed with representative genes of these networks. The model was validated in two test sets. Survival analysis was performed using the output of the model to evaluate whether it could predict patients' survival independent of clinicopathological variables. Five networks were significantly associated with patients' survival. Considering prognostic significance and representativeness, genes of the two survival-related networks were selected for input of the model. The output of the model was significantly associated with patients' survival in two test sets and training set (p < 0.00001, p < 0.0001 and p = 0.02 for training and test sets 1 and 2, resp.). In multivariate analyses, the model was associated with patients' prognosis independent of other clinicopathological features. Our study presents a new perspective on incorporating gene coexpression networks into the gene expression signature and clinical application of deep learning in genomic data science for prognosis prediction. PMID:29581968

Detection of eardrum abnormalities using ensemble deep learning approaches

NASA Astrophysics Data System (ADS)

Senaras, Caglar; Moberly, Aaron C.; Teknos, Theodoros; Essig, Garth; Elmaraghy, Charles; Taj-Schaal, Nazhat; Yua, Lianbo; Gurcan, Metin N.

2018-02-01

In this study, we proposed an approach to report the condition of the eardrum as "normal" or "abnormal" by ensembling two different deep learning architectures. In the first network (Network 1), we applied transfer learning to the Inception V3 network by using 409 labeled samples. As a second network (Network 2), we designed a convolutional neural network to take advantage of auto-encoders by using additional 673 unlabeled eardrum samples. The individual classification accuracies of the Network 1 and Network 2 were calculated as 84.4%(+/- 12.1%) and 82.6% (+/- 11.3%), respectively. Only 32% of the errors of the two networks were the same, making it possible to combine two approaches to achieve better classification accuracy. The proposed ensemble method allows us to achieve robust classification because it has high accuracy (84.4%) with the lowest standard deviation (+/- 10.3%).

Deep neural networks to enable real-time multimessenger astrophysics

NASA Astrophysics Data System (ADS)

George, Daniel; Huerta, E. A.

2018-02-01

Gravitational wave astronomy has set in motion a scientific revolution. To further enhance the science reach of this emergent field of research, there is a pressing need to increase the depth and speed of the algorithms used to enable these ground-breaking discoveries. We introduce Deep Filtering—a new scalable machine learning method for end-to-end time-series signal processing. Deep Filtering is based on deep learning with two deep convolutional neural networks, which are designed for classification and regression, to detect gravitational wave signals in highly noisy time-series data streams and also estimate the parameters of their sources in real time. Acknowledging that some of the most sensitive algorithms for the detection of gravitational waves are based on implementations of matched filtering, and that a matched filter is the optimal linear filter in Gaussian noise, the application of Deep Filtering using whitened signals in Gaussian noise is investigated in this foundational article. The results indicate that Deep Filtering outperforms conventional machine learning techniques, achieves similar performance compared to matched filtering, while being several orders of magnitude faster, allowing real-time signal processing with minimal resources. Furthermore, we demonstrate that Deep Filtering can detect and characterize waveform signals emitted from new classes of eccentric or spin-precessing binary black holes, even when trained with data sets of only quasicircular binary black hole waveforms. The results presented in this article, and the recent use of deep neural networks for the identification of optical transients in telescope data, suggests that deep learning can facilitate real-time searches of gravitational wave sources and their electromagnetic and astroparticle counterparts. In the subsequent article, the framework introduced herein is directly applied to identify and characterize gravitational wave events in real LIGO data.

Deep convolutional neural network for prostate MR segmentation

NASA Astrophysics Data System (ADS)

Tian, Zhiqiang; Liu, Lizhi; Fei, Baowei

2017-03-01

Automatic segmentation of the prostate in magnetic resonance imaging (MRI) has many applications in prostate cancer diagnosis and therapy. We propose a deep fully convolutional neural network (CNN) to segment the prostate automatically. Our deep CNN model is trained end-to-end in a single learning stage based on prostate MR images and the corresponding ground truths, and learns to make inference for pixel-wise segmentation. Experiments were performed on our in-house data set, which contains prostate MR images of 20 patients. The proposed CNN model obtained a mean Dice similarity coefficient of 85.3%+/-3.2% as compared to the manual segmentation. Experimental results show that our deep CNN model could yield satisfactory segmentation of the prostate.

Curriculum Framework (CF) Implementation Conference. Report of the Regional Educational Laboratory Network Program and the National Network of Eisenhower Mathematics and Science Regional Consortia (Hilton Head Island, South Carolina, January 26-27, 1995).

ERIC Educational Resources Information Center

Palmer, Jackie; Powell, Mary Jo

The Laboratory Network Program and the National Network of Eisenhower Mathematics and Science Regional Consortia, operating as the Curriculum Frameworks Task Force, jointly convened a group of educators involved in implementing state-level mathematics or science curriculum frameworks (CF). The Hilton Head (South Carolina) conference had a dual…

Geophysical Observations Supporting Research of Magmatic Processes at Icelandic Volcanoes

NASA Astrophysics Data System (ADS)

Vogfjörd, Kristín. S.; Hjaltadóttir, Sigurlaug; Roberts, Matthew J.

2010-05-01

Magmatic processes at volcanoes on the boundary between the European and North American plates in Iceland are observed with in-situ multidisciplinary geophysical networks owned by different national, European or American universities and research institutions, but through collaboration mostly operated by the Icelandic Meteorological Office. The terrestrial observations are augmented by space-based interferometric synthetic aperture radar (InSAR) images of the volcanoes and their surrounding surface. Together this infrastructure can monitor magma movements in several volcanoes from the base of the crust up to the surface. The national seismic network is sensitive enough to detect small scale seismicity deep in the crust under some of the voclanoes. High resolution mapping of this seismicity and its temporal progression has been used to delineate the track of the magma as it migrates upwards in the crust, either to form an intrusion at shallow levels or to reach the surface in an eruption. Broadband recording has also enabled capturing low frequency signals emanating from magmatic movements. In two volcanoes, Eyjafjallajökull and Katla, just east of the South Iceland Seismic Zone (SISZ), seismicity just above the crust-mantle boundary has revealed magma intruding into the crust from the mantle below. As the magma moves to shallower levels, the deformation of the Earth‘s surface is captured by geodetic systems, such as continuous GPS networks, (InSAR) images of the surface and -- even more sensitive to the deformation -- strain meters placed in boreholes around 200 m below the Earth‘s surface. Analysis of these signals can reveal the size and shape of the magma as well as the temporal evolution. At near-by Hekla volcano flanking the SISZ to the north, where only 50% of events are of M>1 compared to 86% of earthquakes in Eyjafjallajökull, the sensitivity of the seismic network is insufficient to detect the smallest seismicity and so the volcano appears less active and deep seismicity has not been detected. Improved seismic station density may improve the resolution of deep processes. Due do Hekla‘s continued expansion, the concentration of the continuous GPS network has been increased around Hekla and a strain meter will be installed by the volcano in 2010. The increased density of geodetic observations is expected to increase the resolution of the depth, volume and geometry of the magma chamber. Before the volcano's latest eruption in 2000, the increased seismicity and deformation signal recorded by the nearest seismic station and strain meter (at 15 km distance) enabled a public warning to be issued of the impending eruption 30 minutes prior to eruption. The additional instrumentation around Hekla is expected to extend the previous advance warning time.

Characteristics of Offshore Hawai';i Island Seismicity and Velocity Structure, including Lo';ihi Submarine Volcano

NASA Astrophysics Data System (ADS)

Merz, D. K.; Caplan-Auerbach, J.; Thurber, C. H.

2013-12-01

The Island of Hawai';i is home to the most active volcanoes in the Hawaiian Islands. The island's isolated nature, combined with the lack of permanent offshore seismometers, creates difficulties in recording small magnitude earthquakes with accuracy. This background offshore seismicity is crucial in understanding the structure of the lithosphere around the island chain, the stresses on the lithosphere generated by the weight of the islands, and how the volcanoes interact with each other offshore. This study uses the data collected from a 9-month deployment of a temporary ocean bottom seismometer (OBS) network fully surrounding Lo';ihi volcano. This allowed us to widen the aperture of earthquake detection around the Big Island, lower the magnitude detection threshold, and better constrain the hypocentral depths of offshore seismicity that occurs between the OBS network and the Hawaii Volcano Observatory's land based network. Although this study occurred during a time of volcanic quiescence for Lo';ihi, it establishes a basis for background seismicity of the volcano. More than 480 earthquakes were located using the OBS network, incorporating data from the HVO network where possible. Here we present relocated hypocenters using the double-difference earthquake location algorithm HypoDD (Waldhauser & Ellsworth, 2000), as well as tomographic images for a 30 km square area around the summit of Lo';ihi. Illuminated by using the double-difference earthquake location algorithm HypoDD (Waldhauser & Ellsworth, 2000), offshore seismicity during this study is punctuated by events locating in the mantle fault zone 30-50km deep. These events reflect rupture on preexisting faults in the lower lithosphere caused by stresses induced by volcano loading and flexure of the Pacific Plate (Wolfe et al., 2004; Pritchard et al., 2007). Tomography was performed using the double-difference seismic tomography method TomoDD (Zhang & Thurber, 2003) and showed overall velocities to be slower than the regional velocity model (HG50; Klein, 1989) in the shallow lithosphere above 16 km depth. This is likely a result of thick deposits of volcaniclastic sediments and fractured pillow basalts that blanket the southern submarine flank of Mauna Loa, upon which Lo';ihi is currently superimposing (Morgan et al., 2003). A broad, low-velocity anomaly was observed from 20-40 km deep beneath the area of Pahala, and is indicative of the central plume conduit that supplies magma to the active volcanoes. A localized high-velocity body is observed 4-6 km deep beneath Lo';ihi's summit, extending 10 km to the North and South. Oriented approximately parallel to Lo';ihi's active rift zones, this high-velocity body is suggestive of intrusion in the upper crust, similar to Kilauea's high-velocity rift zones.

Automatic Seismic-Event Classification with Convolutional Neural Networks.

NASA Astrophysics Data System (ADS)

Bueno Rodriguez, A.; Titos Luzón, M.; Garcia Martinez, L.; Benitez, C.; Ibáñez, J. M.

2017-12-01

Active volcanoes exhibit a wide range of seismic signals, providing vast amounts of unlabelled volcano-seismic data that can be analyzed through the lens of artificial intelligence. However, obtaining high-quality labelled data is time-consuming and expensive. Deep neural networks can process data in their raw form, compute high-level features and provide a better representation of the input data distribution. These systems can be deployed to classify seismic data at scale, enhance current early-warning systems and build extensive seismic catalogs. In this research, we aim to classify spectrograms from seven different seismic events registered at "Volcán de Fuego" (Colima, Mexico), during four eruptive periods. Our approach is based on convolutional neural networks (CNNs), a sub-type of deep neural networks that can exploit grid structure from the data. Volcano-seismic signals can be mapped into a grid-like structure using the spectrogram: a representation of the temporal evolution in terms of time and frequency. Spectrograms were computed from the data using Hamming windows with 4 seconds length, 2.5 seconds overlapping and 128 points FFT resolution. Results are compared to deep neural networks, random forest and SVMs. Experiments show that CNNs can exploit temporal and frequency information, attaining a classification accuracy of 93%, similar to deep networks 91% but outperforming SVM and random forest. These results empirically show that CNNs are powerful models to classify a wide range of volcano-seismic signals, and achieve good generalization. Furthermore, volcano-seismic spectrograms contains useful discriminative information for the CNN, as higher layers of the network combine high-level features computed for each frequency band, helping to detect simultaneous events in time. Being at the intersection of deep learning and geophysics, this research enables future studies of how CNNs can be used in volcano monitoring to accurately determine the detection and location of seismic events.

Fluxes of Soot Carbon to South Atlantic Sediments

EPA Science Inventory

Deep sea sediment samples from the South Atlantic Ocean were analyzed for soot black carbon (BC), total organic carbon (TOC), stable carbon isotope ratios (δ 13C), and polycyclic aromatic hydrocarbons (PAHs). Soot BC was present at low concentrations (0.04–0.17% dry weight), but ...

Earth's crust model of the South-Okhotsk Basin by wide-angle OBS data

NASA Astrophysics Data System (ADS)

Kashubin, Sergey N.; Petrov, Oleg V.; Rybalka, Alexander V.; Milshtein, Evgenia D.; Shokalsky, Sergey P.; Verba, Mark L.; Petrov, Evgeniy O.

2017-07-01

Deep seismic studies of the Sea of Okhotsk region started in late 1950s. Since that time, wide-angle reflection and refraction data on more than two dozen profiles were acquired. Only five of those profiles either crossed or entered the deep-water area of the South-Okhotsk Basin (also known as the Kuril Basin or the South-Okhotsk Deep-Water Trough). Only P-waves were used to develop velocity-interface models in all the early research. Thus, all seismic and geodynamic models of the Okhotsk region were based only on the information on compressional waves. Nevertheless, the use of Vp/Vs ratio in addition to P-wave velocity allows discriminating felsic and mafic crustal layers with similar Vp values. In 2007 the Russian seismic service company Sevmorgeo acquired multi-component data with ocean bottom seismometers (OBS) along the 1700-km-long north-south 2-DV-M Profile. Only P-wave information was used previously to develop models for the entire profile. In this study, a multi-wave processing, analysis, and interpretation of the OBS data are presented for the 550-km-long southern segment of this Profile that crosses the deep-water South-Okhotsk Basin. Within this segment 50 seismometers were deployed with nominal OBS station spacing of 10-12 km. Shot point spacing was 250 m. Not only primary P-waves and S-waves but also multiples and P-S, S-P converted waves were analyzed in this study to constrain velocity-interface models by means of travel time forward modeling. In offshore deep seismic studies, thick water layer hinders an estimation of velocities in the sedimentary cover and in the upper consolidated crust. Primarily, this is due to the fact that refracted waves propagating in low-velocity solid upper layers interfere with high-amplitude direct water wave. However, in multi-component measurements with ocean bottom seismometers, it is possible to use converted and multiple waves for velocity estimations in these layers. Consequently, one can obtain P- and S-waves velocity models of the sedimentary strata and the upper consolidated crust. Velocity values in the upper consolidated crust beneath the South-Okhotsk Basin (Vp = 5.50-5.80 km/s, Vp/Vs = 1.74-1.76) allow interpretation of this 2.5-3.5-km-thick layer to be consistent with a felsic (granodioritic) crust. These results suggest that the Earth's crust in this region can be considered continental in nature, rather than previously accepted oceanic crust. Even though, the crust is thinned and stretched at this location.

Event-Driven Random Back-Propagation: Enabling Neuromorphic Deep Learning Machines

PubMed Central

Neftci, Emre O.; Augustine, Charles; Paul, Somnath; Detorakis, Georgios

2017-01-01

An ongoing challenge in neuromorphic computing is to devise general and computationally efficient models of inference and learning which are compatible with the spatial and temporal constraints of the brain. One increasingly popular and successful approach is to take inspiration from inference and learning algorithms used in deep neural networks. However, the workhorse of deep learning, the gradient descent Gradient Back Propagation (BP) rule, often relies on the immediate availability of network-wide information stored with high-precision memory during learning, and precise operations that are difficult to realize in neuromorphic hardware. Remarkably, recent work showed that exact backpropagated gradients are not essential for learning deep representations. Building on these results, we demonstrate an event-driven random BP (eRBP) rule that uses an error-modulated synaptic plasticity for learning deep representations. Using a two-compartment Leaky Integrate & Fire (I&F) neuron, the rule requires only one addition and two comparisons for each synaptic weight, making it very suitable for implementation in digital or mixed-signal neuromorphic hardware. Our results show that using eRBP, deep representations are rapidly learned, achieving classification accuracies on permutation invariant datasets comparable to those obtained in artificial neural network simulations on GPUs, while being robust to neural and synaptic state quantizations during learning. PMID:28680387

Event-Driven Random Back-Propagation: Enabling Neuromorphic Deep Learning Machines.

PubMed

Neftci, Emre O; Augustine, Charles; Paul, Somnath; Detorakis, Georgios

2017-01-01

An ongoing challenge in neuromorphic computing is to devise general and computationally efficient models of inference and learning which are compatible with the spatial and temporal constraints of the brain. One increasingly popular and successful approach is to take inspiration from inference and learning algorithms used in deep neural networks. However, the workhorse of deep learning, the gradient descent Gradient Back Propagation (BP) rule, often relies on the immediate availability of network-wide information stored with high-precision memory during learning, and precise operations that are difficult to realize in neuromorphic hardware. Remarkably, recent work showed that exact backpropagated gradients are not essential for learning deep representations. Building on these results, we demonstrate an event-driven random BP (eRBP) rule that uses an error-modulated synaptic plasticity for learning deep representations. Using a two-compartment Leaky Integrate & Fire (I&F) neuron, the rule requires only one addition and two comparisons for each synaptic weight, making it very suitable for implementation in digital or mixed-signal neuromorphic hardware. Our results show that using eRBP, deep representations are rapidly learned, achieving classification accuracies on permutation invariant datasets comparable to those obtained in artificial neural network simulations on GPUs, while being robust to neural and synaptic state quantizations during learning.

Classify epithelium-stroma in histopathological images based on deep transferable network.

PubMed

Yu, X; Zheng, H; Liu, C; Huang, Y; Ding, X

2018-04-20

Recently, the deep learning methods have received more attention in histopathological image analysis. However, the traditional deep learning methods assume that training data and test data have the same distributions, which causes certain limitations in real-world histopathological applications. However, it is costly to recollect a large amount of labeled histology data to train a new neural network for each specified image acquisition procedure even for similar tasks. In this paper, an unsupervised domain adaptation is introduced into a typical deep convolutional neural network (CNN) model to mitigate the repeating of the labels. The unsupervised domain adaptation is implemented by adding two regularisation terms, namely the feature-based adaptation and entropy minimisation, to the object function of a widely used CNN model called the AlexNet. Three independent public epithelium-stroma datasets were used to verify the proposed method. The experimental results have demonstrated that in the epithelium-stroma classification, the proposed method can achieve better performance than the commonly used deep learning methods and some existing deep domain adaptation methods. Therefore, the proposed method can be considered as a better option for the real-world applications of histopathological image analysis because there is no requirement for recollection of large-scale labeled data for every specified domain. © 2018 The Authors Journal of Microscopy © 2018 Royal Microscopical Society.

Genesis of Miocene litho-stratigraphic trap and hydrocarbon accumulation in the Qiongdongnan Basin, northern South China Sea

NASA Astrophysics Data System (ADS)

Fan, Caiwei; Jiang, Tao; Liu, Kun; Tan, Jiancai; Li, Hu; Li, Anqi

2018-12-01

In recent years, several large gas fields have been discovered in western Qiongdongnan Basin. It is important and necessary to illustrate their sedimentary characteristics and hydrocarbon migration so that more gas fields could be discovered in the future. Previous regional tectonic-sedimentary researchers show that large-scale source rock of the Yacheng Formation developed in the Ledong and Lingshui sags due to the Red River Fault pull-apart strike slip in early Oligocene. The main targets for hydrocarbon exploration in this area are the Miocene deep water reservoirs. In late Miocene, the Huangliu Formation reservoirs are composed of the early channels which were sourced by river systems in Hainan uplift and the consequent channels were sourced by Qiupen River in Kunsong uplift. Both axial channels exhibit unique spatial distribution patterns and geometries. The other kind of reservoir developed in the middle Miocene Meishan Formation, which compose of slope break-controlled submarine fan. They can be further classified into three types—slope channelized fan, basin floor fan, and bottom current reworked fan. The various fans have different reservoir quality. These two kinds of reservoirs contribute to four types of litho-stratigraphic traps under the actions of sedimentation and subsidence. The overpressure caused by hydrocarbon generation can fracture deeper strata and result in regional fractured network for hydrocarbon migration. Therefore, free gas driven by overpressure and buoyancy force can be migrated into Miocene litho-stratigraphic traps to accumulate. The revealed genesis of Miocene lithologic trap and hydrocarbon accumulation in the Qiongdongnan Basin would greatly contribute to the further hydrocarbon exploration in northern South China Sea and can be helpful for other deep water areas around the world.

Fabric defect detection based on visual saliency using deep feature and low-rank recovery

NASA Astrophysics Data System (ADS)

Liu, Zhoufeng; Wang, Baorui; Li, Chunlei; Li, Bicao; Dong, Yan

2018-04-01

Fabric defect detection plays an important role in improving the quality of fabric product. In this paper, a novel fabric defect detection method based on visual saliency using deep feature and low-rank recovery was proposed. First, unsupervised training is carried out by the initial network parameters based on MNIST large datasets. The supervised fine-tuning of fabric image library based on Convolutional Neural Networks (CNNs) is implemented, and then more accurate deep neural network model is generated. Second, the fabric images are uniformly divided into the image block with the same size, then we extract their multi-layer deep features using the trained deep network. Thereafter, all the extracted features are concentrated into a feature matrix. Third, low-rank matrix recovery is adopted to divide the feature matrix into the low-rank matrix which indicates the background and the sparse matrix which indicates the salient defect. In the end, the iterative optimal threshold segmentation algorithm is utilized to segment the saliency maps generated by the sparse matrix to locate the fabric defect area. Experimental results demonstrate that the feature extracted by CNN is more suitable for characterizing the fabric texture than the traditional LBP, HOG and other hand-crafted features extraction method, and the proposed method can accurately detect the defect regions of various fabric defects, even for the image with complex texture.

Exemplar-Based Image and Video Stylization Using Fully Convolutional Semantic Features.

PubMed

Zhu, Feida; Yan, Zhicheng; Bu, Jiajun; Yu, Yizhou

2017-05-10

Color and tone stylization in images and videos strives to enhance unique themes with artistic color and tone adjustments. It has a broad range of applications from professional image postprocessing to photo sharing over social networks. Mainstream photo enhancement softwares, such as Adobe Lightroom and Instagram, provide users with predefined styles, which are often hand-crafted through a trial-and-error process. Such photo adjustment tools lack a semantic understanding of image contents and the resulting global color transform limits the range of artistic styles it can represent. On the other hand, stylistic enhancement needs to apply distinct adjustments to various semantic regions. Such an ability enables a broader range of visual styles. In this paper, we first propose a novel deep learning architecture for exemplar-based image stylization, which learns local enhancement styles from image pairs. Our deep learning architecture consists of fully convolutional networks (FCNs) for automatic semantics-aware feature extraction and fully connected neural layers for adjustment prediction. Image stylization can be efficiently accomplished with a single forward pass through our deep network. To extend our deep network from image stylization to video stylization, we exploit temporal superpixels (TSPs) to facilitate the transfer of artistic styles from image exemplars to videos. Experiments on a number of datasets for image stylization as well as a diverse set of video clips demonstrate the effectiveness of our deep learning architecture.

DEEP MOTIF DASHBOARD: VISUALIZING AND UNDERSTANDING GENOMIC SEQUENCES USING DEEP NEURAL NETWORKS.

PubMed

Lanchantin, Jack; Singh, Ritambhara; Wang, Beilun; Qi, Yanjun

2017-01-01

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence's saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them.

Deep Motif Dashboard: Visualizing and Understanding Genomic Sequences Using Deep Neural Networks

PubMed Central

Lanchantin, Jack; Singh, Ritambhara; Wang, Beilun; Qi, Yanjun

2018-01-01

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence’s saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them. PMID:27896980

Chasing a complete understanding of a rapid moving rock slide: the La Saxe landslide

NASA Astrophysics Data System (ADS)

Crosta, G. B.; Cancelli, P.; Tamburini, A.; Alberto, W.; Broccolato, M.; Castellanza, R.; Frattini, P.; Agliardi, F.; Rivolta, C.; Leva, D.

2012-04-01

Large deep seated slope deformations affect entire valley flanks and are characterized by slow to extremely slow present day displacement rates. Because of their extreme size, they are frequently characterized at their interior by secondary instabilities which can be classified as rockslides, that can originate large rock avalanches or can move at much faster rates with respect to the main mass. As a consequence local instabilities and reactivation of sectors of deep seated deformations should be carefully monitored and studied especially because they can affect strongly deformed and weakened rock masses. Because of these natural conditions and their preferential location in coincidence of slope steepening, these rockslides can undergo rapid evolution and activation putting the upmost urgency for monitoring, hazard and risk assessment. We present the case study of the La Saxe rockslide (Courmayeur, Aosta valley, Italy), located within a deep seated deformation affecting most of the 10 km long left hand flank of the Ferret valley (between 1340 m and 2300 m a.s.l.) and which underwent a major phase of acceleration in the last decade. The rockslide affects the extreme south western tip of the deep seated deformation at the outlet of Ferret valley, with an estimated volume of about 8 x 106 m3 of clay schists and thinly bedded black carbonates, intensely folded and faulted. An intense investigation activity has been performed in the last 2 years to reach a more complete understanding of the phenomenon. Boreholes have been drilled, logged, and instrumented to constrain the landslide volume, the rate of displacement at depth, and the water pressure. Displacement monitoring has been undertaken at successive steps by setting up sequentially: a distance measurement network (6 optical targets), a GPS network for periodic measurements (12 stations), a ground-based interferometer (GB-InSAR, LisaLab, by Ellegi, with 10 min acquisition intervals), a geodetic network based on a total station and 25 optical targets measured at 2 h intervals, a GPS network (7 stations) for quasi-real time measurements, four differential multiparametric borehole systems (DMS columns up to 100 m long). A geotechnical network has been also implemented including open pipe piezometers, borehole wire extensometers and inclinometric casings. This enormous monitoring effort is motivated by the extreme risk associated to this phenomenon, which is hanging over a famous touristic resort, a world famous cable way, the Mont Blanc highway, and in close proximity to the Mont Blanc tunnel. Rockslide characterization, failure surface definition, and groundwater flow investigations allowed for a series of slope stability analyses to be completed, together with modelling of the expected invasion area. Relationships with snowmelt have been ascertained and an early warning system based on real time measurements redundancy and all weather capabilities has been set up. LisaLab GB-InSAR equipment continuously provide spatially distributed displacement data which have been analysed to identify different failure scenarios and sensitivity of the landslide to triggering and controlling factors. Geodetic measurements are integrated with GB-InSAR data for verification and in depth 3D displacement reconstructions.

Fusion of shallow and deep features for classification of high-resolution remote sensing images

NASA Astrophysics Data System (ADS)

Gao, Lang; Tian, Tian; Sun, Xiao; Li, Hang

2018-02-01

Effective spectral and spatial pixel description plays a significant role for the classification of high resolution remote sensing images. Current approaches of pixel-based feature extraction are of two main kinds: one includes the widelyused principal component analysis (PCA) and gray level co-occurrence matrix (GLCM) as the representative of the shallow spectral and shape features, and the other refers to the deep learning-based methods which employ deep neural networks and have made great promotion on classification accuracy. However, the former traditional features are insufficient to depict complex distribution of high resolution images, while the deep features demand plenty of samples to train the network otherwise over fitting easily occurs if only limited samples are involved in the training. In view of the above, we propose a GLCM-based convolution neural network (CNN) approach to extract features and implement classification for high resolution remote sensing images. The employment of GLCM is able to represent the original images and eliminate redundant information and undesired noises. Meanwhile, taking shallow features as the input of deep network will contribute to a better guidance and interpretability. In consideration of the amount of samples, some strategies such as L2 regularization and dropout methods are used to prevent over-fitting. The fine-tuning strategy is also used in our study to reduce training time and further enhance the generalization performance of the network. Experiments with popular data sets such as PaviaU data validate that our proposed method leads to a performance improvement compared to individual involved approaches.

The deep space network, volume 10

NASA Technical Reports Server (NTRS)

1972-01-01

Progress on the Deep Space Network (DSN) supporting research and technology is reported. The objectives, functions and facilities of the DSN are described along with the mission support for the following: interplanetary flight projects, planetary flight projects, and manned space flight projects. Work in advanced engineering and communications systems is reported along with changes in hardware and software configurations in the DSN/MSFN tracking stations.

A Spatiotemporal Prediction Framework for Air Pollution Based on Deep RNN

NASA Astrophysics Data System (ADS)

Fan, J.; Li, Q.; Hou, J.; Feng, X.; Karimian, H.; Lin, S.

2017-10-01

Time series data in practical applications always contain missing values due to sensor malfunction, network failure, outliers etc. In order to handle missing values in time series, as well as the lack of considering temporal properties in machine learning models, we propose a spatiotemporal prediction framework based on missing value processing algorithms and deep recurrent neural network (DRNN). By using missing tag and missing interval to represent time series patterns, we implement three different missing value fixing algorithms, which are further incorporated into deep neural network that consists of LSTM (Long Short-term Memory) layers and fully connected layers. Real-world air quality and meteorological datasets (Jingjinji area, China) are used for model training and testing. Deep feed forward neural networks (DFNN) and gradient boosting decision trees (GBDT) are trained as baseline models against the proposed DRNN. Performances of three missing value fixing algorithms, as well as different machine learning models are evaluated and analysed. Experiments show that the proposed DRNN framework outperforms both DFNN and GBDT, therefore validating the capacity of the proposed framework. Our results also provides useful insights for better understanding of different strategies that handle missing values.

Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition.

PubMed

Ordóñez, Francisco Javier; Roggen, Daniel

2016-01-18

Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters' influence on performance to provide insights about their optimisation.

DRREP: deep ridge regressed epitope predictor.

PubMed

Sher, Gene; Zhi, Degui; Zhang, Shaojie

2017-10-03

The ability to predict epitopes plays an enormous role in vaccine development in terms of our ability to zero in on where to do a more thorough in-vivo analysis of the protein in question. Though for the past decade there have been numerous advancements and improvements in epitope prediction, on average the best benchmark prediction accuracies are still only around 60%. New machine learning algorithms have arisen within the domain of deep learning, text mining, and convolutional networks. This paper presents a novel analytically trained and string kernel using deep neural network, which is tailored for continuous epitope prediction, called: Deep Ridge Regressed Epitope Predictor (DRREP). DRREP was tested on long protein sequences from the following datasets: SARS, Pellequer, HIV, AntiJen, and SEQ194. DRREP was compared to numerous state of the art epitope predictors, including the most recently published predictors called LBtope and DMNLBE. Using area under ROC curve (AUC), DRREP achieved a performance improvement over the best performing predictors on SARS (13.7%), HIV (8.9%), Pellequer (1.5%), and SEQ194 (3.1%), with its performance being matched only on the AntiJen dataset, by the LBtope predictor, where both DRREP and LBtope achieved an AUC of 0.702. DRREP is an analytically trained deep neural network, thus capable of learning in a single step through regression. By combining the features of deep learning, string kernels, and convolutional networks, the system is able to perform residue-by-residue prediction of continues epitopes with higher accuracy than the current state of the art predictors.

Deep hierarchical attention network for video description

NASA Astrophysics Data System (ADS)

Li, Shuohao; Tang, Min; Zhang, Jun

2018-03-01

Pairing video to natural language description remains a challenge in computer vision and machine translation. Inspired by image description, which uses an encoder-decoder model for reducing visual scene into a single sentence, we propose a deep hierarchical attention network for video description. The proposed model uses convolutional neural network (CNN) and bidirectional LSTM network as encoders while a hierarchical attention network is used as the decoder. Compared to encoder-decoder models used in video description, the bidirectional LSTM network can capture the temporal structure among video frames. Moreover, the hierarchical attention network has an advantage over single-layer attention network on global context modeling. To make a fair comparison with other methods, we evaluate the proposed architecture with different types of CNN structures and decoders. Experimental results on the standard datasets show that our model has a more superior performance than the state-of-the-art techniques.

A model-guided symbolic execution approach for network protocol implementations and vulnerability detection.

PubMed

Wen, Shameng; Meng, Qingkun; Feng, Chao; Tang, Chaojing

2017-01-01

Formal techniques have been devoted to analyzing whether network protocol specifications violate security policies; however, these methods cannot detect vulnerabilities in the implementations of the network protocols themselves. Symbolic execution can be used to analyze the paths of the network protocol implementations, but for stateful network protocols, it is difficult to reach the deep states of the protocol. This paper proposes a novel model-guided approach to detect vulnerabilities in network protocol implementations. Our method first abstracts a finite state machine (FSM) model, then utilizes the model to guide the symbolic execution. This approach achieves high coverage of both the code and the protocol states. The proposed method is implemented and applied to test numerous real-world network protocol implementations. The experimental results indicate that the proposed method is more effective than traditional fuzzing methods such as SPIKE at detecting vulnerabilities in the deep states of network protocol implementations.

Subsurface and terrain controls on runoff generation in deep soil landscapes

NASA Astrophysics Data System (ADS)

Mallard, John; McGlynn, Brian; Richter, Daniel

2017-04-01

Our understanding of runoff generation in regions characterized by deep, highly weathered soils is incomplete despite the prevalence of this setting worldwide. To address this, we instrumented a first-order watershed in the Piedmont of South Carolina, USA. The Piedmont region of the United States extends east of the Appalachians from Maryland to Alabama, and is home to some of the most rapid population growth in the country. Regional and local relief is modest, although the landscape is highly dissected and local slope can be quite variable. The region's soils are ancient, deeply weathered, and characterized by sharp changes in hydrologic properties due to concentration of clay in the Bt horizon. Despite a mild climate and consistent precipitation, seasonally variable energy availability and deciduous tree cover create a strong evapotranspiration mediated seasonal hydrologic dynamic: while moist soils and extended stream networks are typical of the late fall through spring, relatively dry soils and contracting stream networks emerge in the summer and early fall. To elucidate the control of the complex vertical and planform structure of this region, as well as the strongly seasonal subsurface hydrology, on runoff generation, we installed a network of nested, shallow groundwater wells across an ephemeral to first-order watershed to continuously measure internal water levels. We also recorded local precipitation and discharge at the outlet of this watershed, a similar adjacent watershed, and in the second to third order downstream watershed. Subsurface water dynamics varied spatially, vertically, and seasonally. Shallow depths and landscape positions with minimal contributing area exhibited flashier dynamics comparable to the stream hydrographs while positions with more contributing area exhibited relatively muted dynamics. Most well positions showed minimal response to precipitation throughout the summer, and even occasionally observed response rarely co-occurred with streamflow generation. Our initial findings suggest that characterizing the terrain of a watershed must be coupled with the subsurface soil hydrology in order to understand spatiotemporal patterns of streamflow generation in regions possessing both complex vertical structure and terrain.

Terrain and subsurface influences on runoff generation in a steep, deep, highly weathered system

NASA Astrophysics Data System (ADS)

Mallard, J. M.; McGlynn, B. L.; Richter, D. D., Jr.

2017-12-01

Our understanding of runoff generation in regions characterized by deep, highly weathered soils is incomplete, despite the prevalence occupation of these landscapes worldwide. To address this, we instrumented a first-order watershed in the Piedmont of South Carolina, USA, a region that extends east of the Appalachians from Maryland to Alabama, and home to some of the most rapid population growth in the country. Although regionally the relief is modest, the landscape is often highly dissected and local slopes can be steep and highly varied. The typical soils of the region are kaolinite dominated ultisols, with hydrologic properties controlled by argillic Bt horizons, often with >50% clay-size fraction. The humid subtropical climate creates relatively consistent precipitation intra-annually and seasonally variable energy availability. Consequently, the mixed deciduous and coniferous tree cover creates a strong evapotranspiration-mediated hydrologic dynamic. While moist soils and extended stream networks are typical from late fall through spring, relatively dry soils and contracting stream networks emerge in the summer and early fall. Here, we seek to elucidate the relative influence of the vertical soil and spatial terrain structure of this region on watershed hillslope hydrology and subsequent runoff generation. We installed a network of nested, shallow groundwater wells and soil water content probes within an ephemeral to first-order watershed to continuously measure soil and groundwater dynamics across soil horizons and landscape position. We also recorded local precipitation and discharge from this watershed. Most landscape positions exhibited minimal water table response to precipitation throughout dry summer periods, with infrequently observed responses rarely coincident with streamflow generation. In contrast, during the wetter late fall through early spring period, streamflow was driven by the interaction between transient perched water tables and topographically mediated redistribution of shallow groundwater downslope. Our findings suggest that understanding streamflow generation in regions possessing both complex terrain and complex vertical soil structure requires synchronous characterization of terrain mediated water redistribution and subsurface soil hydrology.

The Ties that Bind: Race and Restitution in Education Law and Policy in South Africa and the United States of America

ERIC Educational Resources Information Center

Jansen, Jonathan D.

2006-01-01

The parallels between South Africa and the United States run deep. For the United States, that moment of transition, at least as far as education is concerned, was the landmark ruling of 1954, described in the shorthand, "Brown v. Board of Education"; for South Africa, that moment came 40 years later when every citizen could, for the…

Infrared Faint Radio Sources in the Extended Chandra Deep Field South

NASA Astrophysics Data System (ADS)

Huynh, Minh T.

2009-01-01

Infrared-Faint Radio Sources (IFRSs) are a class of radio objects found in the Australia Telescope Large Area Survey (ATLAS) which have no observable counterpart in the Spitzer Wide-area Infrared Extragalactic Survey (SWIRE). The extended Chandra Deep Field South now has even deeper Spitzer imaging (3.6 to 70 micron) from a number of Legacy surveys. We report the detections of two IFRS sources in IRAC images. The non-detection of two other IFRSs allows us to constrain the source type. Detailed modeling of the SED of these objects shows that they are consistent with high redshift AGN (z > 2).

Global latitudinal species diversity gradient in deep-sea benthic foraminifera

NASA Astrophysics Data System (ADS)

Culver, Stephen J.; Buzas, Martin A.

2000-02-01

Global scale patterns of species diversity for modern deep-sea benthic foraminifera, an important component of the bathyal and abyssal meiofauna, are examined using comparable data from five studies in the Atlantic, ranging over 138° of latitude from the Norwegian Sea to the Weddell Sea. We show that a pattern of decreasing diversity with increasing latitude characterises both the North and South Atlantic. This pattern is confirmed for the northern hemisphere by independent data from the west-central North Atlantic and the Arctic basin. Species diversity in the North Atlantic northwards from the equator is variable until a sharp fall in the Norwegian Sea (ca. 65°N). In the South Atlantic species diversity drops from a maximum in latitudes less than 30°S and then decreases slightly from 40 to 70°S. For any given latitude, North Atlantic diversity is generally lower than in the South Atlantic. Both ecological and historical factors related to food supply are invoked to explain the formation and maintenance of the latitudinal gradient of deep-sea benthic foraminiferal species diversity. The gradient formed some 36 million years ago when global climatic cooling led to seasonally fluctuating food supply in higher latitudes.

Reducing Cancer Disparities Through Innovative Partnerships: A Collaboration of the South Carolina Cancer Prevention and Control Research Network and Federally Qualified Health Centers

PubMed Central

Young, Vicki M.; Freedman, Darcy A.; Adams, Swann Arp; Brandt, Heather M.; Xirasagar, Sudha; Felder, Tisha M.; Ureda, John R.; Hurley, Thomas; Khang, Leepao; Campbell, Dayna; Hébert, James R.

2011-01-01

The South Carolina Cancer Prevention and Control Research Network, in partnership with the South Carolina Primary Health Care Association, and Federally Qualified Health Centers (FQHCs), aims to promote evidence-based cancer interventions in community-based primary care settings. Partnership activities include (1) examining FQHCs’ readiness and capacity for conducting research, (2) developing a cancer-focused data sharing network, and (3) integrating a farmers’ market within an FQHC. These activities identify unique opportunities for public health and primary care collaborations. PMID:21932143

Deep convolutional neural networks for estimating porous material parameters with ultrasound tomography

NASA Astrophysics Data System (ADS)

Lähivaara, Timo; Kärkkäinen, Leo; Huttunen, Janne M. J.; Hesthaven, Jan S.

2018-02-01

We study the feasibility of data based machine learning applied to ultrasound tomography to estimate water-saturated porous material parameters. In this work, the data to train the neural networks is simulated by solving wave propagation in coupled poroviscoelastic-viscoelastic-acoustic media. As the forward model, we consider a high-order discontinuous Galerkin method while deep convolutional neural networks are used to solve the parameter estimation problem. In the numerical experiment, we estimate the material porosity and tortuosity while the remaining parameters which are of less interest are successfully marginalized in the neural networks-based inversion. Computational examples confirms the feasibility and accuracy of this approach.

Self-Organized Information Processing in Neuronal Networks: Replacing Layers in Deep Networks by Dynamics

NASA Astrophysics Data System (ADS)

Kirst, Christoph

It is astonishing how the sub-parts of a brain co-act to produce coherent behavior. What are mechanism that coordinate information processing and communication and how can those be changed flexibly in order to cope with variable contexts? Here we show that when information is encoded in the deviations around a collective dynamical reference state of a recurrent network the propagation of these fluctuations is strongly dependent on precisely this underlying reference. Information here 'surfs' on top of the collective dynamics and switching between states enables fast and flexible rerouting of information. This in turn affects local processing and consequently changes in the global reference dynamics that re-regulate the distribution of information. This provides a generic mechanism for self-organized information processing as we demonstrate with an oscillatory Hopfield network that performs contextual pattern recognition. Deep neural networks have proven to be very successful recently. Here we show that generating information channels via collective reference dynamics can effectively compress a deep multi-layer architecture into a single layer making this mechanism a promising candidate for the organization of information processing in biological neuronal networks.

Cascaded deep decision networks for classification of endoscopic images

NASA Astrophysics Data System (ADS)

Murthy, Venkatesh N.; Singh, Vivek; Sun, Shanhui; Bhattacharya, Subhabrata; Chen, Terrence; Comaniciu, Dorin

2017-02-01

Both traditional and wireless capsule endoscopes can generate tens of thousands of images for each patient. It is desirable to have the majority of irrelevant images filtered out by automatic algorithms during an offline review process or to have automatic indication for highly suspicious areas during an online guidance. This also applies to the newly invented endomicroscopy, where online indication of tumor classification plays a significant role. Image classification is a standard pattern recognition problem and is well studied in the literature. However, performance on the challenging endoscopic images still has room for improvement. In this paper, we present a novel Cascaded Deep Decision Network (CDDN) to improve image classification performance over standard Deep neural network based methods. During the learning phase, CDDN automatically builds a network which discards samples that are classified with high confidence scores by a previously trained network and concentrates only on the challenging samples which would be handled by the subsequent expert shallow networks. We validate CDDN using two different types of endoscopic imaging, which includes a polyp classification dataset and a tumor classification dataset. From both datasets we show that CDDN can outperform other methods by about 10%. In addition, CDDN can also be applied to other image classification problems.

Strain accumulation in southern California, 1973-1980.

USGS Publications Warehouse

Savage, J.C.; Prescott, W.H.; Lisowski, M.; King, N.E.

1981-01-01

Frequent surveys of seven trilateration networks in southern California over the interval 1973-1980 suggest that a regional increment in strain may have occurred in 1978-1979. Prior to 1978 and after late 1979 the strain accumulation has been predominantly a uniaxial north-south compression. This secular trend was interrupted sometime in 1978-1979 by an increment in both north-south and east-west extension in five of the seven networks. The onset of this change appears to have occurred first in the networks farthest south. The changes occurred without any unusual seismicity within the networks, but the overall seismicity in southern California was unusually low prior to and has been unusually high since the occurrence. The average principal strain rates for the seven networks in the 1973-1980 interval are 0.17 mu strain/yr north- south contraction and 0.08 mu strain/yr east-west extension. Although the observed increment in strain could be related to unidentified systematic error in the measuring system, a careful review of the measurements and comparisons with three other measuring systems reveal no appreciable cumulative systematic error. -Authors

The Network Concept of Creativity and Deep Thinking: Applications to Social Opinion Formation and Talent Support

ERIC Educational Resources Information Center

Csermely, Peter

2017-01-01

Our century has unprecedented new challenges, which need creative solutions and deep thinking. Contemplative, deep thinking became an "endangered species" in our rushing world of Tweets, elevator pitches, and fast decisions. Here, we describe how important aspects of both creativity and deep thinking can be understood as network…

Neutrino oscillation studies with IceCube-DeepCore

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

Aartsen, M. G.; Abraham, K.; Ackermann, M.

IceCube, a gigaton-scale neutrino detector located at the South Pole, was primarily designed to search for astrophysical neutrinos with energies of PeV and higher. This goal has been achieved with the detection of the highest energy neutrinos to date. At the other end of the energy spectrum, the DeepCore extension lowers the energy threshold of the detector to approximately 10 GeV and opens the door for oscillation studies using atmospheric neutrinos. An analysis of the disappearance of these neutrinos has been completed, with the results produced being complementary with dedicated oscillation experiments. Following a review of the detector principle andmore » performance, the method used to make these calculations, as well as the results, is detailed. Finally, the future prospects of IceCube-DeepCore and the next generation of neutrino experiments at the South Pole (IceCube-Gen2, specifically the PINGU sub-detector) are briefly discussed.« less

Neutrino oscillation studies with IceCube-DeepCore

DOE PAGES

Aartsen, M. G.; Abraham, K.; Ackermann, M.; ...

2016-03-30

IceCube, a gigaton-scale neutrino detector located at the South Pole, was primarily designed to search for astrophysical neutrinos with energies of PeV and higher. This goal has been achieved with the detection of the highest energy neutrinos to date. At the other end of the energy spectrum, the DeepCore extension lowers the energy threshold of the detector to approximately 10 GeV and opens the door for oscillation studies using atmospheric neutrinos. An analysis of the disappearance of these neutrinos has been completed, with the results produced being complementary with dedicated oscillation experiments. Following a review of the detector principle andmore » performance, the method used to make these calculations, as well as the results, is detailed. Finally, the future prospects of IceCube-DeepCore and the next generation of neutrino experiments at the South Pole (IceCube-Gen2, specifically the PINGU sub-detector) are briefly discussed.« less

A history of the deep space network

NASA Technical Reports Server (NTRS)

Corliss, W. R.

1976-01-01

The Deep Space Network (DSN) has been managed and operated by the Jet Propulsion Laboratory (JPL) under NASA contract ever since NASA was formed in late 1958. The Tracking and data acquisition tasks of the DSN are markedly different from those of the other NASA network, STDN. STDN, which is an amalgamation of the satellite tracking network (STADAN) and the Manned Space Flight Network (MSFN), is primarily concerned with supporting manned and unmanned earth satellites. In contrast, the DSN deals with spacecraft that are thousands to hundreds of millions of miles away. The radio signals from these distant craft are many orders of magnitude weaker than those from nearby satellites. Distance also makes precise radio location more difficult; and accurate trajectory data are vital to deep space navigation in the vicinities of the other planets of the solar system. In addition to tracking spacecraft and acquiring data from them, the DSN is required to transmit many thousands of commands to control the sophisticated planetary probes and interplanetary monitoring stations. To meet these demanding requirements, the DSN has been compelled to be in the forefront of technology.

Feature to prototype transition in neural networks

NASA Astrophysics Data System (ADS)

Krotov, Dmitry; Hopfield, John

Models of associative memory with higher order (higher than quadratic) interactions, and their relationship to neural networks used in deep learning are discussed. Associative memory is conventionally described by recurrent neural networks with dynamical convergence to stable points. Deep learning typically uses feedforward neural nets without dynamics. However, a simple duality relates these two different views when applied to problems of pattern classification. From the perspective of associative memory such models deserve attention because they make it possible to store a much larger number of memories, compared to the quadratic case. In the dual description, these models correspond to feedforward neural networks with one hidden layer and unusual activation functions transmitting the activities of the visible neurons to the hidden layer. These activation functions are rectified polynomials of a higher degree rather than the rectified linear functions used in deep learning. The network learns representations of the data in terms of features for rectified linear functions, but as the power in the activation function is increased there is a gradual shift to a prototype-based representation, the two extreme regimes of pattern recognition known in cognitive psychology. Simons Center for Systems Biology.

Robust visual tracking via multiscale deep sparse networks

NASA Astrophysics Data System (ADS)

Wang, Xin; Hou, Zhiqiang; Yu, Wangsheng; Xue, Yang; Jin, Zefenfen; Dai, Bo

2017-04-01

In visual tracking, deep learning with offline pretraining can extract more intrinsic and robust features. It has significant success solving the tracking drift in a complicated environment. However, offline pretraining requires numerous auxiliary training datasets and is considerably time-consuming for tracking tasks. To solve these problems, a multiscale sparse networks-based tracker (MSNT) under the particle filter framework is proposed. Based on the stacked sparse autoencoders and rectifier linear unit, the tracker has a flexible and adjustable architecture without the offline pretraining process and exploits the robust and powerful features effectively only through online training of limited labeled data. Meanwhile, the tracker builds four deep sparse networks of different scales, according to the target's profile type. During tracking, the tracker selects the matched tracking network adaptively in accordance with the initial target's profile type. It preserves the inherent structural information more efficiently than the single-scale networks. Additionally, a corresponding update strategy is proposed to improve the robustness of the tracker. Extensive experimental results on a large scale benchmark dataset show that the proposed method performs favorably against state-of-the-art methods in challenging environments.

Overview of deep learning in medical imaging.

PubMed

Suzuki, Kenji

2017-09-01

The use of machine learning (ML) has been increasing rapidly in the medical imaging field, including computer-aided diagnosis (CAD), radiomics, and medical image analysis. Recently, an ML area called deep learning emerged in the computer vision field and became very popular in many fields. It started from an event in late 2012, when a deep-learning approach based on a convolutional neural network (CNN) won an overwhelming victory in the best-known worldwide computer vision competition, ImageNet Classification. Since then, researchers in virtually all fields, including medical imaging, have started actively participating in the explosively growing field of deep learning. In this paper, the area of deep learning in medical imaging is overviewed, including (1) what was changed in machine learning before and after the introduction of deep learning, (2) what is the source of the power of deep learning, (3) two major deep-learning models: a massive-training artificial neural network (MTANN) and a convolutional neural network (CNN), (4) similarities and differences between the two models, and (5) their applications to medical imaging. This review shows that ML with feature input (or feature-based ML) was dominant before the introduction of deep learning, and that the major and essential difference between ML before and after deep learning is the learning of image data directly without object segmentation or feature extraction; thus, it is the source of the power of deep learning, although the depth of the model is an important attribute. The class of ML with image input (or image-based ML) including deep learning has a long history, but recently gained popularity due to the use of the new terminology, deep learning. There are two major models in this class of ML in medical imaging, MTANN and CNN, which have similarities as well as several differences. In our experience, MTANNs were substantially more efficient in their development, had a higher performance, and required a lesser number of training cases than did CNNs. "Deep learning", or ML with image input, in medical imaging is an explosively growing, promising field. It is expected that ML with image input will be the mainstream area in the field of medical imaging in the next few decades.

Conversion of Continuous-Valued Deep Networks to Efficient Event-Driven Networks for Image Classification.

PubMed

Rueckauer, Bodo; Lungu, Iulia-Alexandra; Hu, Yuhuang; Pfeiffer, Michael; Liu, Shih-Chii

2017-01-01

Spiking neural networks (SNNs) can potentially offer an efficient way of doing inference because the neurons in the networks are sparsely activated and computations are event-driven. Previous work showed that simple continuous-valued deep Convolutional Neural Networks (CNNs) can be converted into accurate spiking equivalents. These networks did not include certain common operations such as max-pooling, softmax, batch-normalization and Inception-modules. This paper presents spiking equivalents of these operations therefore allowing conversion of nearly arbitrary CNN architectures. We show conversion of popular CNN architectures, including VGG-16 and Inception-v3, into SNNs that produce the best results reported to date on MNIST, CIFAR-10 and the challenging ImageNet dataset. SNNs can trade off classification error rate against the number of available operations whereas deep continuous-valued neural networks require a fixed number of operations to achieve their classification error rate. From the examples of LeNet for MNIST and BinaryNet for CIFAR-10, we show that with an increase in error rate of a few percentage points, the SNNs can achieve more than 2x reductions in operations compared to the original CNNs. This highlights the potential of SNNs in particular when deployed on power-efficient neuromorphic spiking neuron chips, for use in embedded applications.

Conversion of Continuous-Valued Deep Networks to Efficient Event-Driven Networks for Image Classification

PubMed Central

Rueckauer, Bodo; Lungu, Iulia-Alexandra; Hu, Yuhuang; Pfeiffer, Michael; Liu, Shih-Chii

2017-01-01

Spiking neural networks (SNNs) can potentially offer an efficient way of doing inference because the neurons in the networks are sparsely activated and computations are event-driven. Previous work showed that simple continuous-valued deep Convolutional Neural Networks (CNNs) can be converted into accurate spiking equivalents. These networks did not include certain common operations such as max-pooling, softmax, batch-normalization and Inception-modules. This paper presents spiking equivalents of these operations therefore allowing conversion of nearly arbitrary CNN architectures. We show conversion of popular CNN architectures, including VGG-16 and Inception-v3, into SNNs that produce the best results reported to date on MNIST, CIFAR-10 and the challenging ImageNet dataset. SNNs can trade off classification error rate against the number of available operations whereas deep continuous-valued neural networks require a fixed number of operations to achieve their classification error rate. From the examples of LeNet for MNIST and BinaryNet for CIFAR-10, we show that with an increase in error rate of a few percentage points, the SNNs can achieve more than 2x reductions in operations compared to the original CNNs. This highlights the potential of SNNs in particular when deployed on power-efficient neuromorphic spiking neuron chips, for use in embedded applications. PMID:29375284

The Deep Impact Network Experiment Operations Center

NASA Technical Reports Server (NTRS)

Torgerson, J. Leigh; Clare, Loren; Wang, Shin-Ywan

2009-01-01

Delay/Disruption Tolerant Networking (DTN) promises solutions in solving space communications challenges arising from disconnections as orbiters lose line-of-sight with landers, long propagation delays over interplanetary links, and other phenomena. DTN has been identified as the basis for the future NASA space communications network backbone, and international standardization is progressing through both the Consultative Committee for Space Data Systems (CCSDS) and the Internet Engineering Task Force (IETF). JPL has developed an implementation of the DTN architecture, called the Interplanetary Overlay Network (ION). ION is specifically implemented for space use, including design for use in a real-time operating system environment and high processing efficiency. In order to raise the Technology Readiness Level of ION, the first deep space flight demonstration of DTN is underway, using the Deep Impact (DI) spacecraft. Called the Deep Impact Network (DINET), operations are planned for Fall 2008. An essential component of the DINET project is the Experiment Operations Center (EOC), which will generate and receive the test communications traffic as well as "out-of-DTN band" command and control of the DTN experiment, store DTN flight test information in a database, provide display systems for monitoring DTN operations status and statistics (e.g., bundle throughput), and support query and analyses of the data collected. This paper describes the DINET EOC and its value in the DTN flight experiment and potential for further DTN testing.

OVERVIEW OF AERIAL TRAM SUPPORT TOWERS NINE, TEN, AND DEEP ...

Library of Congress Historic Buildings Survey, Historic Engineering Record, Historic Landscapes Survey

OVERVIEW OF AERIAL TRAM SUPPORT TOWERS NINE, TEN, AND DEEP RAVINE,LOOKING SOUTH FROM BREAK OVER TOWER LOCATION. A SINGLE ORE BUCKET HANGS FROM THE CABLE AT CENTER. DEATH VALLEY'S FLOOR IS IN THE DISTANCE (TOP). - Keane Wonder Mine, Park Route 4 (Daylight Pass Cutoff), Death Valley Junction, Inyo County, CA

Searching for exoplanets using artificial intelligence

NASA Astrophysics Data System (ADS)

Pearson, Kyle A.; Palafox, Leon; Griffith, Caitlin A.

2018-02-01

In the last decade, over a million stars were monitored to detect transiting planets. Manual interpretation of potential exoplanet candidates is labor intensive and subject to human error, the results of which are difficult to quantify. Here we present a new method of detecting exoplanet candidates in large planetary search projects which, unlike current methods uses a neural network. Neural networks, also called "deep learning" or "deep nets" are designed to give a computer perception into a specific problem by training it to recognize patterns. Unlike past transit detection algorithms deep nets learn to recognize planet features instead of relying on hand-coded metrics that humans perceive as the most representative. Our convolutional neural network is capable of detecting Earth-like exoplanets in noisy time-series data with a greater accuracy than a least-squares method. Deep nets are highly generalizable allowing data to be evaluated from different time series after interpolation without compromising performance. As validated by our deep net analysis of Kepler light curves, we detect periodic transits consistent with the true period without any model fitting. Our study indicates that machine learning will facilitate the characterization of exoplanets in future analysis of large astronomy data sets.

viral abundance distribution in deep waters of the Northern of South China Sea

NASA Astrophysics Data System (ADS)

He, Lei; Yin, Kedong

2017-04-01

Little is known about the vertical distribution and interaction of viruses and bacteria in the deep ocean water column. The vertical distribution of viral-like particles and bacterial abundance was investigated in the deep water column in the South China Sea during September 2005 along with salinity, temperature and dissolved oxygen. There were double maxima in the ratio of viral to bacterial abundance (VBR) in the water column: the subsurface maximum located at 50-100 m near the pycnocline layer, and the deep maximum at 800-1000 m. At the subsurface maximum of VBR, both viral and bacterial abundance were maximal in the water column, and at the deep maximum of VBR, both viral and bacterial abundance were low, but bacterial abundance was relatively lower than viral abundance. The subsurface VBR maximum coincided with the subsurface chlorophyll maximum while the deep VBR maximum coincided with the minimum in dissolved oxygen (2.91mg L-1). Therefore, we hypothesize that the two maxima were formed by different mechanisms. The subsurface VBR maximum was formed due to an increase in bacterial abundance resulting from the stimulation of abundant organic supply at the subsurface chlorophyll maximum, whereas the deep VBR maximum was formed due to a decrease in bacterial abundance caused by more limitation of organic matter at the oxygen minimum. The evidence suggests that viruses play an important role in controlling bacterial abundance in the deep water column due to the limitation of organic matter supply. In turn, this slows down the formation of the oxygen minimum in which oxygen may be otherwise lower. The mechanism has a great implication that viruses could control bacterial decomposition of organic matter, oxygen consumption and nutrient remineralization in the deep oceans.

Deep Restricted Kernel Machines Using Conjugate Feature Duality.

PubMed

Suykens, Johan A K

2017-08-01

The aim of this letter is to propose a theory of deep restricted kernel machines offering new foundations for deep learning with kernel machines. From the viewpoint of deep learning, it is partially related to restricted Boltzmann machines, which are characterized by visible and hidden units in a bipartite graph without hidden-to-hidden connections and deep learning extensions as deep belief networks and deep Boltzmann machines. From the viewpoint of kernel machines, it includes least squares support vector machines for classification and regression, kernel principal component analysis (PCA), matrix singular value decomposition, and Parzen-type models. A key element is to first characterize these kernel machines in terms of so-called conjugate feature duality, yielding a representation with visible and hidden units. It is shown how this is related to the energy form in restricted Boltzmann machines, with continuous variables in a nonprobabilistic setting. In this new framework of so-called restricted kernel machine (RKM) representations, the dual variables correspond to hidden features. Deep RKM are obtained by coupling the RKMs. The method is illustrated for deep RKM, consisting of three levels with a least squares support vector machine regression level and two kernel PCA levels. In its primal form also deep feedforward neural networks can be trained within this framework.

South-Central Tibetan Seismicity from HiCLIMB Seismic Array Data

NASA Astrophysics Data System (ADS)

Carpenter, S.; Nabelek, J.; Braunmiller, J.

2010-12-01

The HiCLIMB broadband passive seismic experiment (2002-2005) operated 233 sites along a 800-km long north-south array extending from the Himalayan foreland into the Central Tibetan Plateau and a flanking 350x350 km lateral array in southern Tibet and eastern Nepal. We use data from the experiment’s second phase (June 2004 to August 2005), when stations operated in Tibet, to locate earthquakes in south-central Tibet, a region with no permanent seismic network where little is known about its seismicity. We used the Antelope software for automatic detection and arrival time picking, event-arrival association and event location. Requiring a low detection and event association threshold initially resulted in ~110,000 declared events. The large database size rendered manual inspection unfeasible and we developed automated post-processing modules to weed out spurious detections and erroneous phase and event associations, which stemmed, e.g., from multiple coincident earthquakes within the array or misplaced seismicity from the great 2004 Sumatra earthquake. The resulting database contains ~32,000 events within 5° distance from the closest station. We consider ~7,600 events defined by more than 30 P and S arrivals well located and discuss them here. Seismicity in the subset correlates well with mapped faults and structures seen on satellite imagery attesting to high location quality. This is confirmed by non-systematic, kilometer-scale differences between automatic and manual locations for selected events. Seismicity in south-central Tibet is intense north of the Yarlung-Tsangpo Suture. Almost 90% of events occurred in the Lhasa Terrane mainly along north-south trending rifts. Vigorous activity (>4,800 events) accompanied two M>6 earthquakes in the Payang Basin (84°E), ~100 km west of the linear array. The Tangra-Yum Co (86.5°E) and Pumqu-Xianza (88°E) rifts were very active (~1,000 events) without dominant main shocks indicating swarm like-behavior possibly related to shallow magmatic or geothermal activity. Seismicity in the Qiangtang Terrane accounts for less than 10% of activity; seismicity is distributed and, except for the Yibuk-Caka Rift (87°E), difficult to associate with known structures. Lower seismicity may be apparent and simply reflect a larger distance to the array. Fewer than 5% of events occurred south of the Yarlong Tsangpo Suture in the Tethyan Himalaya, the only region where in addition to shallow seismicity a significant number of deep (mantle) events was located. Hypocenter depth, particularly for shallow events, is usually not well constrained due to array geometry and large distances to closest sites. The nature of deep events inside the array, though, is resolved.

Pubface: Celebrity face identification based on deep learning

NASA Astrophysics Data System (ADS)

Ouanan, H.; Ouanan, M.; Aksasse, B.

2018-05-01

In this paper, we describe a new real time application called PubFace, which allows to recognize celebrities in public spaces by employs a new pose invariant face recognition deep neural network algorithm with an extremely low error rate. To build this application, we make the following contributions: firstly, we build a novel dataset with over five million faces labelled. Secondly, we fine tuning the deep convolutional neural network (CNN) VGG-16 architecture on our new dataset that we have built. Finally, we deploy this model on the Raspberry Pi 3 model B using the OpenCv dnn module (OpenCV 3.3).

Deep learning on temporal-spectral data for anomaly detection

NASA Astrophysics Data System (ADS)

Ma, King; Leung, Henry; Jalilian, Ehsan; Huang, Daniel

2017-05-01

Detecting anomalies is important for continuous monitoring of sensor systems. One significant challenge is to use sensor data and autonomously detect changes that cause different conditions to occur. Using deep learning methods, we are able to monitor and detect changes as a result of some disturbance in the system. We utilize deep neural networks for sequence analysis of time series. We use a multi-step method for anomaly detection. We train the network to learn spectral and temporal features from the acoustic time series. We test our method using fiber-optic acoustic data from a pipeline.

The deep-sea as a final global sink of semivolatile persistent organic pollutants? Part I: PCBs in surface and deep-sea dwelling fish of the north and south Atlantic and the Monterey Bay Canyon (California).

PubMed

Froescheis, O; Looser, R; Cailliet, G M; Jarman, W M; Ballschmiter, K

2000-03-01

The understanding of the global environmental multiphase distribution of persistent organic pollutants (POPs) as a result of the physico-chemical properties of the respective compounds is well established. We have analysed the results of a vertical transport of POPs from upper water layers (0-200 m) to the deepwater region (> 800 m) in terms of the contamination of the biophase in both water layers. The contents of persistent organochlorine compounds like polychlorinated biphenyls (PCBs) in fish living in the upper water layers of the North Atlantic and the South Atlantic, and at the continental shelf of California (Marine Sanctuary Monterey Bay and its deep-sea Canyon) are compared to the levels in deep-sea or bottom dwelling fish within the same geographic area. The deep-sea biota show significantly higher burdens as compared to surface-living species of the same region. There are also indications for recycling processes of POPs--in this case the PCBs--in the biophase of the abyss as well. It can be concluded that the bio- and geo phase of the deep-sea may act similarly as the upper horizons of forest and grasslands on the continents as an ultimate global sink for POPs in the marine environment.

Applying the Uses and Gratifications Theory to Compare Higher Education Students' Motivation for Using Social Networking Sites: Experiences from Iran, Malaysia, United Kingdom, and South Africa

ERIC Educational Resources Information Center

Karimi, Leila; Khodabandelou, Rouhollah; Ehsani, Maryam; Ahmad, Muhammad

2014-01-01

Drawing from the Uses and Gratifications Theory, this study examined the Gratification Sought and the Gratification Obtained from using Social Networking Sites among Iranian, Malaysian, British, and South African higher education students. This comparison allowed to drawing conclusions about how social networking sites fulfill users' needs with…

Joint Segmentation of Multiple Thoracic Organs in CT Images with Two Collaborative Deep Architectures.

PubMed

Trullo, Roger; Petitjean, Caroline; Nie, Dong; Shen, Dinggang; Ruan, Su

2017-09-01

Computed Tomography (CT) is the standard imaging technique for radiotherapy planning. The delineation of Organs at Risk (OAR) in thoracic CT images is a necessary step before radiotherapy, for preventing irradiation of healthy organs. However, due to low contrast, multi-organ segmentation is a challenge. In this paper, we focus on developing a novel framework for automatic delineation of OARs. Different from previous works in OAR segmentation where each organ is segmented separately, we propose two collaborative deep architectures to jointly segment all organs, including esophagus, heart, aorta and trachea. Since most of the organ borders are ill-defined, we believe spatial relationships must be taken into account to overcome the lack of contrast. The aim of combining two networks is to learn anatomical constraints with the first network, which will be used in the second network, when each OAR is segmented in turn. Specifically, we use the first deep architecture, a deep SharpMask architecture, for providing an effective combination of low-level representations with deep high-level features, and then take into account the spatial relationships between organs by the use of Conditional Random Fields (CRF). Next, the second deep architecture is employed to refine the segmentation of each organ by using the maps obtained on the first deep architecture to learn anatomical constraints for guiding and refining the segmentations. Experimental results show superior performance on 30 CT scans, comparing with other state-of-the-art methods.

Incorporating deep learning with convolutional neural networks and position specific scoring matrices for identifying electron transport proteins.

PubMed

Le, Nguyen-Quoc-Khanh; Ho, Quang-Thai; Ou, Yu-Yen

2017-09-05

In several years, deep learning is a modern machine learning technique using in a variety of fields with state-of-the-art performance. Therefore, utilization of deep learning to enhance performance is also an important solution for current bioinformatics field. In this study, we try to use deep learning via convolutional neural networks and position specific scoring matrices to identify electron transport proteins, which is an important molecular function in transmembrane proteins. Our deep learning method can approach a precise model for identifying of electron transport proteins with achieved sensitivity of 80.3%, specificity of 94.4%, and accuracy of 92.3%, with MCC of 0.71 for independent dataset. The proposed technique can serve as a powerful tool for identifying electron transport proteins and can help biologists understand the function of the electron transport proteins. Moreover, this study provides a basis for further research that can enrich a field of applying deep learning in bioinformatics. © 2017 Wiley Periodicals, Inc. © 2017 Wiley Periodicals, Inc.

Deep learning architecture for iris recognition based on optimal Gabor filters and deep belief network

NASA Astrophysics Data System (ADS)

He, Fei; Han, Ye; Wang, Han; Ji, Jinchao; Liu, Yuanning; Ma, Zhiqiang

2017-03-01

Gabor filters are widely utilized to detect iris texture information in several state-of-the-art iris recognition systems. However, the proper Gabor kernels and the generative pattern of iris Gabor features need to be predetermined in application. The traditional empirical Gabor filters and shallow iris encoding ways are incapable of dealing with such complex variations in iris imaging including illumination, aging, deformation, and device variations. Thereby, an adaptive Gabor filter selection strategy and deep learning architecture are presented. We first employ particle swarm optimization approach and its binary version to define a set of data-driven Gabor kernels for fitting the most informative filtering bands, and then capture complex pattern from the optimal Gabor filtered coefficients by a trained deep belief network. A succession of comparative experiments validate that our optimal Gabor filters may produce more distinctive Gabor coefficients and our iris deep representations be more robust and stable than traditional iris Gabor codes. Furthermore, the depth and scales of the deep learning architecture are also discussed.

pDeep: Predicting MS/MS Spectra of Peptides with Deep Learning.

PubMed

Zhou, Xie-Xuan; Zeng, Wen-Feng; Chi, Hao; Luo, Chunjie; Liu, Chao; Zhan, Jianfeng; He, Si-Min; Zhang, Zhifei

2017-12-05

In tandem mass spectrometry (MS/MS)-based proteomics, search engines rely on comparison between an experimental MS/MS spectrum and the theoretical spectra of the candidate peptides. Hence, accurate prediction of the theoretical spectra of peptides appears to be particularly important. Here, we present pDeep, a deep neural network-based model for the spectrum prediction of peptides. Using the bidirectional long short-term memory (BiLSTM), pDeep can predict higher-energy collisional dissociation, electron-transfer dissociation, and electron-transfer and higher-energy collision dissociation MS/MS spectra of peptides with >0.9 median Pearson correlation coefficients. Further, we showed that intermediate layer of the neural network could reveal physicochemical properties of amino acids, for example the similarities of fragmentation behaviors between amino acids. We also showed the potential of pDeep to distinguish extremely similar peptides (peptides that contain isobaric amino acids, for example, GG = N, AG = Q, or even I = L), which were very difficult to distinguish using traditional search engines.

Uncertain Photometric Redshifts with Deep Learning Methods

NASA Astrophysics Data System (ADS)

D'Isanto, A.

2017-06-01

The need for accurate photometric redshifts estimation is a topic that has fundamental importance in Astronomy, due to the necessity of efficiently obtaining redshift information without the need of spectroscopic analysis. We propose a method for determining accurate multi-modal photo-z probability density functions (PDFs) using Mixture Density Networks (MDN) and Deep Convolutional Networks (DCN). A comparison with a Random Forest (RF) is performed.

Nonparametric Representations for Integrated Inference, Control, and Sensing

DTIC Science & Technology

2015-10-01

Learning (ICML), 2013. [20] Jeff Donahue, Yangqing Jia, Oriol Vinyals, Judy Hoffman, Ning Zhang, Eric Tzeng, and Trevor Darrell. DeCAF: A deep ...unlimited. Multi-layer feature learning “SuperVision” Convolutional Neural Network (CNN) ImageNet Classification with Deep Convolutional Neural Networks...to develop a new framework for autonomous operations that will extend the state of the art in distributed learning and modeling from data, and

Simple gain probability functions for large reflector antennas of JPL/NASA

NASA Technical Reports Server (NTRS)

Jamnejad, V.

2003-01-01

Simple models for the patterns as well as their cumulative gain probability and probability density functions of the Deep Space Network antennas are developed. These are needed for the study and evaluation of interference from unwanted sources such as the emerging terrestrial system, High Density Fixed Service, with the Ka-band receiving antenna systems in Goldstone Station of the Deep Space Network.

The Telecommunications and Data Acquisition Report. [Deep Space Network

NASA Technical Reports Server (NTRS)

Posner, E. C. (Editor)

1988-01-01

In space communications, radio navigation, radio science, and ground based radio and radar astronomy, activities of the Deep Space Network and its associated Ground Communications Facility in planning, in supporting research and technology, in implementation, and in operations are reported. Also included is TDA funded activity at JPL on data and information systems and reimbursable DSN work performed for other space agencies through NASA.

Beyond Retinal Layers: A Deep Voting Model for Automated Geographic Atrophy Segmentation in SD-OCT Images

PubMed Central

Ji, Zexuan; Chen, Qiang; Niu, Sijie; Leng, Theodore; Rubin, Daniel L.

2018-01-01

Purpose To automatically and accurately segment geographic atrophy (GA) in spectral-domain optical coherence tomography (SD-OCT) images by constructing a voting system with deep neural networks without the use of retinal layer segmentation. Methods An automatic GA segmentation method for SD-OCT images based on the deep network was constructed. The structure of the deep network was composed of five layers, including one input layer, three hidden layers, and one output layer. During the training phase, the labeled A-scans with 1024 features were directly fed into the network as the input layer to obtain the deep representations. Then a soft-max classifier was trained to determine the label of each individual pixel. Finally, a voting decision strategy was used to refine the segmentation results among 10 trained models. Results Two image data sets with GA were used to evaluate the model. For the first dataset, our algorithm obtained a mean overlap ratio (OR) 86.94% ± 8.75%, absolute area difference (AAD) 11.49% ± 11.50%, and correlation coefficients (CC) 0.9857; for the second dataset, the mean OR, AAD, and CC of the proposed method were 81.66% ± 10.93%, 8.30% ± 9.09%, and 0.9952, respectively. The proposed algorithm was capable of improving over 5% and 10% segmentation accuracy, respectively, when compared with several state-of-the-art algorithms on two data sets. Conclusions Without retinal layer segmentation, the proposed algorithm could produce higher segmentation accuracy and was more stable when compared with state-of-the-art methods that relied on retinal layer segmentation results. Our model may provide reliable GA segmentations from SD-OCT images and be useful in the clinical diagnosis of advanced nonexudative AMD. Translational Relevance Based on the deep neural networks, this study presents an accurate GA segmentation method for SD-OCT images without using any retinal layer segmentation results, and may contribute to improved understanding of advanced nonexudative AMD. PMID:29302382

Beyond Retinal Layers: A Deep Voting Model for Automated Geographic Atrophy Segmentation in SD-OCT Images.

PubMed

Ji, Zexuan; Chen, Qiang; Niu, Sijie; Leng, Theodore; Rubin, Daniel L

2018-01-01

To automatically and accurately segment geographic atrophy (GA) in spectral-domain optical coherence tomography (SD-OCT) images by constructing a voting system with deep neural networks without the use of retinal layer segmentation. An automatic GA segmentation method for SD-OCT images based on the deep network was constructed. The structure of the deep network was composed of five layers, including one input layer, three hidden layers, and one output layer. During the training phase, the labeled A-scans with 1024 features were directly fed into the network as the input layer to obtain the deep representations. Then a soft-max classifier was trained to determine the label of each individual pixel. Finally, a voting decision strategy was used to refine the segmentation results among 10 trained models. Two image data sets with GA were used to evaluate the model. For the first dataset, our algorithm obtained a mean overlap ratio (OR) 86.94% ± 8.75%, absolute area difference (AAD) 11.49% ± 11.50%, and correlation coefficients (CC) 0.9857; for the second dataset, the mean OR, AAD, and CC of the proposed method were 81.66% ± 10.93%, 8.30% ± 9.09%, and 0.9952, respectively. The proposed algorithm was capable of improving over 5% and 10% segmentation accuracy, respectively, when compared with several state-of-the-art algorithms on two data sets. Without retinal layer segmentation, the proposed algorithm could produce higher segmentation accuracy and was more stable when compared with state-of-the-art methods that relied on retinal layer segmentation results. Our model may provide reliable GA segmentations from SD-OCT images and be useful in the clinical diagnosis of advanced nonexudative AMD. Based on the deep neural networks, this study presents an accurate GA segmentation method for SD-OCT images without using any retinal layer segmentation results, and may contribute to improved understanding of advanced nonexudative AMD.

Bacterial Diversity in the South Adriatic Sea during a Strong, Deep Winter Convection Year

PubMed Central

Korlević, M.; Pop Ristova, P.; Garić, R.; Amann, R.

2014-01-01

The South Adriatic Sea is the deepest part of the Adriatic Sea and represents a key area for both the Adriatic Sea and the deep eastern Mediterranean. It has a role in dense water formation for the eastern Mediterranean deep circulation cell, and it represents an entry point for water masses originating from the Ionian Sea. The biodiversity and seasonality of bacterial picoplankton before, during, and after deep winter convection in the oligotrophic South Adriatic waters were assessed by combining comparative 16S rRNA sequence analysis and catalyzed reporter deposition-fluorescence in situ hybridization (CARD-FISH). The picoplankton communities reached their maximum abundance in the spring euphotic zone when the maximum value of the chlorophyll a in response to deep winter convection was recorded. The communities were dominated by Bacteria, while Archaea were a minor constituent. A seasonality of bacterial richness and diversity was observed, with minimum values occurring during the winter convection and spring postconvection periods and maximum values occurring under summer stratified conditions. The SAR11 clade was the main constituent of the bacterial communities and reached the maximum abundance in the euphotic zone in spring after the convection episode. Cyanobacteria were the second most abundant group, and their abundance strongly depended on the convection event, when minimal cyanobacterial abundance was observed. In spring and autumn, the euphotic zone was characterized by Bacteroidetes and Gammaproteobacteria. Bacteroidetes clades NS2b, NS4, and NS5 and the gammaproteobacterial SAR86 clade were detected to co-occur with phytoplankton blooms. The SAR324, SAR202, and SAR406 clades were present in the deep layer, exhibiting different seasonal variations in abundance. Overall, our data demonstrate that the abundances of particular bacterial clades and the overall bacterial richness and diversity are greatly impacted by strong winter convection. PMID:25548042

Active semi-supervised learning method with hybrid deep belief networks.

PubMed

Zhou, Shusen; Chen, Qingcai; Wang, Xiaolong

2014-01-01

In this paper, we develop a novel semi-supervised learning algorithm called active hybrid deep belief networks (AHD), to address the semi-supervised sentiment classification problem with deep learning. First, we construct the previous several hidden layers using restricted Boltzmann machines (RBM), which can reduce the dimension and abstract the information of the reviews quickly. Second, we construct the following hidden layers using convolutional restricted Boltzmann machines (CRBM), which can abstract the information of reviews effectively. Third, the constructed deep architecture is fine-tuned by gradient-descent based supervised learning with an exponential loss function. Finally, active learning method is combined based on the proposed deep architecture. We did several experiments on five sentiment classification datasets, and show that AHD is competitive with previous semi-supervised learning algorithm. Experiments are also conducted to verify the effectiveness of our proposed method with different number of labeled reviews and unlabeled reviews respectively.

Deep Learning of Orthographic Representations in Baboons

PubMed Central

Hannagan, Thomas; Ziegler, Johannes C.; Dufau, Stéphane; Fagot, Joël; Grainger, Jonathan

2014-01-01

What is the origin of our ability to learn orthographic knowledge? We use deep convolutional networks to emulate the primate's ventral visual stream and explore the recent finding that baboons can be trained to discriminate English words from nonwords [1]. The networks were exposed to the exact same sequence of stimuli and reinforcement signals as the baboons in the experiment, and learned to map real visual inputs (pixels) of letter strings onto binary word/nonword responses. We show that the networks' highest levels of representations were indeed sensitive to letter combinations as postulated in our previous research. The model also captured the key empirical findings, such as generalization to novel words, along with some intriguing inter-individual differences. The present work shows the merits of deep learning networks that can simulate the whole processing chain all the way from the visual input to the response while allowing researchers to analyze the complex representations that emerge during the learning process. PMID:24416300

Detection of bars in galaxies using a deep convolutional neural network

NASA Astrophysics Data System (ADS)

Abraham, Sheelu; Aniyan, A. K.; Kembhavi, Ajit K.; Philip, N. S.; Vaghmare, Kaustubh

2018-06-01

We present an automated method for the detection of bar structure in optical images of galaxies using a deep convolutional neural network that is easy to use and provides good accuracy. In our study, we use a sample of 9346 galaxies in the redshift range of 0.009-0.2 from the Sloan Digital Sky Survey (SDSS), which has 3864 barred galaxies, the rest being unbarred. We reach a top precision of 94 per cent in identifying bars in galaxies using the trained network. This accuracy matches the accuracy reached by human experts on the same data without additional information about the images. Since deep convolutional neural networks can be scaled to handle large volumes of data, the method is expected to have great relevance in an era where astronomy data is rapidly increasing in terms of volume, variety, volatility, and velocity along with other V's that characterize big data. With the trained model, we have constructed a catalogue of barred galaxies from SDSS and made it available online.

Human fatigue expression recognition through image-based dynamic multi-information and bimodal deep learning

NASA Astrophysics Data System (ADS)

Zhao, Lei; Wang, Zengcai; Wang, Xiaojin; Qi, Yazhou; Liu, Qing; Zhang, Guoxin

2016-09-01

Human fatigue is an important cause of traffic accidents. To improve the safety of transportation, we propose, in this paper, a framework for fatigue expression recognition using image-based facial dynamic multi-information and a bimodal deep neural network. First, the landmark of face region and the texture of eye region, which complement each other in fatigue expression recognition, are extracted from facial image sequences captured by a single camera. Then, two stacked autoencoder neural networks are trained for landmark and texture, respectively. Finally, the two trained neural networks are combined by learning a joint layer on top of them to construct a bimodal deep neural network. The model can be used to extract a unified representation that fuses landmark and texture modalities together and classify fatigue expressions accurately. The proposed system is tested on a human fatigue dataset obtained from an actual driving environment. The experimental results demonstrate that the proposed method performs stably and robustly, and that the average accuracy achieves 96.2%.

A light and faster regional convolutional neural network for object detection in optical remote sensing images

NASA Astrophysics Data System (ADS)

Ding, Peng; Zhang, Ye; Deng, Wei-Jian; Jia, Ping; Kuijper, Arjan

2018-07-01

Detection of objects from satellite optical remote sensing images is very important for many commercial and governmental applications. With the development of deep convolutional neural networks (deep CNNs), the field of object detection has seen tremendous advances. Currently, objects in satellite remote sensing images can be detected using deep CNNs. In general, optical remote sensing images contain many dense and small objects, and the use of the original Faster Regional CNN framework does not yield a suitably high precision. Therefore, after careful analysis we adopt dense convoluted networks, a multi-scale representation and various combinations of improvement schemes to enhance the structure of the base VGG16-Net for improving the precision. We propose an approach to reduce the test-time (detection time) and memory requirements. To validate the effectiveness of our approach, we perform experiments using satellite remote sensing image datasets of aircraft and automobiles. The results show that the improved network structure can detect objects in satellite optical remote sensing images more accurately and efficiently.

DeepFix: A Fully Convolutional Neural Network for Predicting Human Eye Fixations.

PubMed

Kruthiventi, Srinivas S S; Ayush, Kumar; Babu, R Venkatesh

2017-09-01

Understanding and predicting the human visual attention mechanism is an active area of research in the fields of neuroscience and computer vision. In this paper, we propose DeepFix, a fully convolutional neural network, which models the bottom-up mechanism of visual attention via saliency prediction. Unlike classical works, which characterize the saliency map using various hand-crafted features, our model automatically learns features in a hierarchical fashion and predicts the saliency map in an end-to-end manner. DeepFix is designed to capture semantics at multiple scales while taking global context into account, by using network layers with very large receptive fields. Generally, fully convolutional nets are spatially invariant-this prevents them from modeling location-dependent patterns (e.g., centre-bias). Our network handles this by incorporating a novel location-biased convolutional layer. We evaluate our model on multiple challenging saliency data sets and show that it achieves the state-of-the-art results.

Image inpainting and super-resolution using non-local recursive deep convolutional network with skip connections

NASA Astrophysics Data System (ADS)

Liu, Miaofeng

2017-07-01

In recent years, deep convolutional neural networks come into use in image inpainting and super-resolution in many fields. Distinct to most of the former methods requiring to know beforehand the local information for corrupted pixels, we propose a 20-depth fully convolutional network to learn an end-to-end mapping a dataset of damaged/ground truth subimage pairs realizing non-local blind inpainting and super-resolution. As there often exist image with huge corruptions or inpainting on a low-resolution image that the existing approaches unable to perform well, we also share parameters in local area of layers to achieve spatial recursion and enlarge the receptive field. To avoid the difficulty of training this deep neural network, skip-connections between symmetric convolutional layers are designed. Experimental results shows that the proposed method outperforms state-of-the-art methods for diverse corrupting and low-resolution conditions, it works excellently when realizing super-resolution and image inpainting simultaneously

Deep Space Network-Wide Portal Development: Planning Service Pilot Project

NASA Technical Reports Server (NTRS)

Doneva, Silviya

2011-01-01

The Deep Space Network (DSN) is an international network of antennas that supports interplanetary spacecraft missions and radio and radar astronomy observations for the exploration of the solar system and the universe. DSN provides the vital two-way communications link that guides and controls planetary explorers, and brings back the images and new scientific information they collect. In an attempt to streamline operations and improve overall services provided by the Deep Space Network a DSN-wide portal is under development. The project is one step in a larger effort to centralize the data collected from current missions including user input parameters for spacecraft to be tracked. This information will be placed into a principal repository where all operations related to the DSN are stored. Furthermore, providing statistical characterization of data volumes will help identify technically feasible tracking opportunities and more precise mission planning by providing upfront scheduling proposals. Business intelligence tools are to be incorporated in the output to deliver data visualization.

The Role of Architectural and Learning Constraints in Neural Network Models: A Case Study on Visual Space Coding.

PubMed

Testolin, Alberto; De Filippo De Grazia, Michele; Zorzi, Marco

2017-01-01

The recent "deep learning revolution" in artificial neural networks had strong impact and widespread deployment for engineering applications, but the use of deep learning for neurocomputational modeling has been so far limited. In this article we argue that unsupervised deep learning represents an important step forward for improving neurocomputational models of perception and cognition, because it emphasizes the role of generative learning as opposed to discriminative (supervised) learning. As a case study, we present a series of simulations investigating the emergence of neural coding of visual space for sensorimotor transformations. We compare different network architectures commonly used as building blocks for unsupervised deep learning by systematically testing the type of receptive fields and gain modulation developed by the hidden neurons. In particular, we compare Restricted Boltzmann Machines (RBMs), which are stochastic, generative networks with bidirectional connections trained using contrastive divergence, with autoencoders, which are deterministic networks trained using error backpropagation. For both learning architectures we also explore the role of sparse coding, which has been identified as a fundamental principle of neural computation. The unsupervised models are then compared with supervised, feed-forward networks that learn an explicit mapping between different spatial reference frames. Our simulations show that both architectural and learning constraints strongly influenced the emergent coding of visual space in terms of distribution of tuning functions at the level of single neurons. Unsupervised models, and particularly RBMs, were found to more closely adhere to neurophysiological data from single-cell recordings in the primate parietal cortex. These results provide new insights into how basic properties of artificial neural networks might be relevant for modeling neural information processing in biological systems.

The Role of Architectural and Learning Constraints in Neural Network Models: A Case Study on Visual Space Coding

PubMed Central

Testolin, Alberto; De Filippo De Grazia, Michele; Zorzi, Marco

2017-01-01

The recent “deep learning revolution” in artificial neural networks had strong impact and widespread deployment for engineering applications, but the use of deep learning for neurocomputational modeling has been so far limited. In this article we argue that unsupervised deep learning represents an important step forward for improving neurocomputational models of perception and cognition, because it emphasizes the role of generative learning as opposed to discriminative (supervised) learning. As a case study, we present a series of simulations investigating the emergence of neural coding of visual space for sensorimotor transformations. We compare different network architectures commonly used as building blocks for unsupervised deep learning by systematically testing the type of receptive fields and gain modulation developed by the hidden neurons. In particular, we compare Restricted Boltzmann Machines (RBMs), which are stochastic, generative networks with bidirectional connections trained using contrastive divergence, with autoencoders, which are deterministic networks trained using error backpropagation. For both learning architectures we also explore the role of sparse coding, which has been identified as a fundamental principle of neural computation. The unsupervised models are then compared with supervised, feed-forward networks that learn an explicit mapping between different spatial reference frames. Our simulations show that both architectural and learning constraints strongly influenced the emergent coding of visual space in terms of distribution of tuning functions at the level of single neurons. Unsupervised models, and particularly RBMs, were found to more closely adhere to neurophysiological data from single-cell recordings in the primate parietal cortex. These results provide new insights into how basic properties of artificial neural networks might be relevant for modeling neural information processing in biological systems. PMID:28377709

A social network typology and sexual risk-taking among men who have sex with men in Cape Town and Port Elizabeth, South Africa

PubMed Central

de Voux, Alex; Baral, Stefan; Bekker, Linda-Gail; Beyrer, Chris; Phaswana-Mafuya, Nancy; Siegler, Aaron; Sullivan, Patrick; Winskell, Kate; Stephenson, Rob

2016-01-01

Despite the high prevalence of HIV among men who have sex with men in South Africa, very little is known about their lived realities, including their social and sexual networks. Given the influence of social network structure on sexual risk behaviours, a better understanding of the social contexts of men who have sex with men is essential for informing the design of HIV programming and messaging. This study explored social network connectivity, an understudied network attribute, examining self-reported connectivity between friends, family and sex partners. Data were collected in Cape Town and Port Elizabeth, South Africa from 78 men who have sex with men who participated in in-depth interviews which included a social network mapping component. Five social network types emerged from the content analysis of these social network maps based on the level of connectivity between family, friends and sex partners, and ranged from disconnected to densely connected networks. The ways in which participants reported sexual risk-taking differed across the five network types revealing diversity in social network profiles. HIV programming and messaging for this population can greatly benefit from recognising the diversity in lived realities and social connections between men who have sex with men. PMID:26569376

A social network typology and sexual risk-taking among men who have sex with men in Cape Town and Port Elizabeth, South Africa.

PubMed

de Voux, Alex; Baral, Stefan D; Bekker, Linda-Gail; Beyrer, Chris; Phaswana-Mafuya, Nancy; Siegler, Aaron J; Sullivan, Patrick S; Winskell, Kate; Stephenson, Rob

2016-01-01

Despite the high prevalence of HIV among men who have sex with men in South Africa, very little is known about their lived realities, including their social and sexual networks. Given the influence of social network structure on sexual risk behaviours, a better understanding of the social contexts of men who have sex with men is essential for informing the design of HIV programming and messaging. This study explored social network connectivity, an understudied network attribute, examining self-reported connectivity between friends, family and sex partners. Data were collected in Cape Town and Port Elizabeth, South Africa, from 78 men who have sex with men who participated in in-depth interviews that included a social network mapping component. Five social network types emerged from the content analysis of these social network maps based on the level of connectivity between family, friends and sex partners, and ranged from disconnected to densely connected networks. The ways in which participants reported sexual risk-taking differed across the five network types, revealing diversity in social network profiles. HIV programming and messaging for this population can greatly benefit from recognising the diversity in lived realities and social connections between men who have sex with men.

Video Salient Object Detection via Fully Convolutional Networks.

PubMed

Wang, Wenguan; Shen, Jianbing; Shao, Ling

This paper proposes a deep learning model to efficiently detect salient regions in videos. It addresses two important issues: 1) deep video saliency model training with the absence of sufficiently large and pixel-wise annotated video data and 2) fast video saliency training and detection. The proposed deep video saliency network consists of two modules, for capturing the spatial and temporal saliency information, respectively. The dynamic saliency model, explicitly incorporating saliency estimates from the static saliency model, directly produces spatiotemporal saliency inference without time-consuming optical flow computation. We further propose a novel data augmentation technique that simulates video training data from existing annotated image data sets, which enables our network to learn diverse saliency information and prevents overfitting with the limited number of training videos. Leveraging our synthetic video data (150K video sequences) and real videos, our deep video saliency model successfully learns both spatial and temporal saliency cues, thus producing accurate spatiotemporal saliency estimate. We advance the state-of-the-art on the densely annotated video segmentation data set (MAE of .06) and the Freiburg-Berkeley Motion Segmentation data set (MAE of .07), and do so with much improved speed (2 fps with all steps).This paper proposes a deep learning model to efficiently detect salient regions in videos. It addresses two important issues: 1) deep video saliency model training with the absence of sufficiently large and pixel-wise annotated video data and 2) fast video saliency training and detection. The proposed deep video saliency network consists of two modules, for capturing the spatial and temporal saliency information, respectively. The dynamic saliency model, explicitly incorporating saliency estimates from the static saliency model, directly produces spatiotemporal saliency inference without time-consuming optical flow computation. We further propose a novel data augmentation technique that simulates video training data from existing annotated image data sets, which enables our network to learn diverse saliency information and prevents overfitting with the limited number of training videos. Leveraging our synthetic video data (150K video sequences) and real videos, our deep video saliency model successfully learns both spatial and temporal saliency cues, thus producing accurate spatiotemporal saliency estimate. We advance the state-of-the-art on the densely annotated video segmentation data set (MAE of .06) and the Freiburg-Berkeley Motion Segmentation data set (MAE of .07), and do so with much improved speed (2 fps with all steps).

Beyond the hype: deep neural networks outperform established methods using a ChEMBL bioactivity benchmark set.

PubMed

Lenselink, Eelke B; Ten Dijke, Niels; Bongers, Brandon; Papadatos, George; van Vlijmen, Herman W T; Kowalczyk, Wojtek; IJzerman, Adriaan P; van Westen, Gerard J P

2017-08-14

The increase of publicly available bioactivity data in recent years has fueled and catalyzed research in chemogenomics, data mining, and modeling approaches. As a direct result, over the past few years a multitude of different methods have been reported and evaluated, such as target fishing, nearest neighbor similarity-based methods, and Quantitative Structure Activity Relationship (QSAR)-based protocols. However, such studies are typically conducted on different datasets, using different validation strategies, and different metrics. In this study, different methods were compared using one single standardized dataset obtained from ChEMBL, which is made available to the public, using standardized metrics (BEDROC and Matthews Correlation Coefficient). Specifically, the performance of Naïve Bayes, Random Forests, Support Vector Machines, Logistic Regression, and Deep Neural Networks was assessed using QSAR and proteochemometric (PCM) methods. All methods were validated using both a random split validation and a temporal validation, with the latter being a more realistic benchmark of expected prospective execution. Deep Neural Networks are the top performing classifiers, highlighting the added value of Deep Neural Networks over other more conventional methods. Moreover, the best method ('DNN_PCM') performed significantly better at almost one standard deviation higher than the mean performance. Furthermore, Multi-task and PCM implementations were shown to improve performance over single task Deep Neural Networks. Conversely, target prediction performed almost two standard deviations under the mean performance. Random Forests, Support Vector Machines, and Logistic Regression performed around mean performance. Finally, using an ensemble of DNNs, alongside additional tuning, enhanced the relative performance by another 27% (compared with unoptimized 'DNN_PCM'). Here, a standardized set to test and evaluate different machine learning algorithms in the context of multi-task learning is offered by providing the data and the protocols. Graphical Abstract .

Trans-species learning of cellular signaling systems with bimodal deep belief networks

PubMed Central

Chen, Lujia; Cai, Chunhui; Chen, Vicky; Lu, Xinghua

2015-01-01

Motivation: Model organisms play critical roles in biomedical research of human diseases and drug development. An imperative task is to translate information/knowledge acquired from model organisms to humans. In this study, we address a trans-species learning problem: predicting human cell responses to diverse stimuli, based on the responses of rat cells treated with the same stimuli. Results: We hypothesized that rat and human cells share a common signal-encoding mechanism but employ different proteins to transmit signals, and we developed a bimodal deep belief network and a semi-restricted bimodal deep belief network to represent the common encoding mechanism and perform trans-species learning. These ‘deep learning’ models include hierarchically organized latent variables capable of capturing the statistical structures in the observed proteomic data in a distributed fashion. The results show that the models significantly outperform two current state-of-the-art classification algorithms. Our study demonstrated the potential of using deep hierarchical models to simulate cellular signaling systems. Availability and implementation: The software is available at the following URL: http://pubreview.dbmi.pitt.edu/TransSpeciesDeepLearning/. The data are available through SBV IMPROVER website, https://www.sbvimprover.com/challenge-2/overview, upon publication of the report by the organizers. Contact: xinghua@pitt.edu Supplementary information: Supplementary data are available at Bioinformatics online. PMID:25995230

Trans-species learning of cellular signaling systems with bimodal deep belief networks.

PubMed

Chen, Lujia; Cai, Chunhui; Chen, Vicky; Lu, Xinghua

2015-09-15

Model organisms play critical roles in biomedical research of human diseases and drug development. An imperative task is to translate information/knowledge acquired from model organisms to humans. In this study, we address a trans-species learning problem: predicting human cell responses to diverse stimuli, based on the responses of rat cells treated with the same stimuli. We hypothesized that rat and human cells share a common signal-encoding mechanism but employ different proteins to transmit signals, and we developed a bimodal deep belief network and a semi-restricted bimodal deep belief network to represent the common encoding mechanism and perform trans-species learning. These 'deep learning' models include hierarchically organized latent variables capable of capturing the statistical structures in the observed proteomic data in a distributed fashion. The results show that the models significantly outperform two current state-of-the-art classification algorithms. Our study demonstrated the potential of using deep hierarchical models to simulate cellular signaling systems. The software is available at the following URL: http://pubreview.dbmi.pitt.edu/TransSpeciesDeepLearning/. The data are available through SBV IMPROVER website, https://www.sbvimprover.com/challenge-2/overview, upon publication of the report by the organizers. xinghua@pitt.edu Supplementary data are available at Bioinformatics online. © The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

Towards Scalable Deep Learning via I/O Analysis and Optimization

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

Pumma, Sarunya; Si, Min; Feng, Wu-Chun

Deep learning systems have been growing in prominence as a way to automatically characterize objects, trends, and anomalies. Given the importance of deep learning systems, researchers have been investigating techniques to optimize such systems. An area of particular interest has been using large supercomputing systems to quickly generate effective deep learning networks: a phase often referred to as “training” of the deep learning neural network. As we scale existing deep learning frameworks—such as Caffe—on these large supercomputing systems, we notice that the parallelism can help improve the computation tremendously, leaving data I/O as the major bottleneck limiting the overall systemmore » scalability. In this paper, we first present a detailed analysis of the performance bottlenecks of Caffe on large supercomputing systems. Our analysis shows that the I/O subsystem of Caffe—LMDB—relies on memory-mapped I/O to access its database, which can be highly inefficient on large-scale systems because of its interaction with the process scheduling system and the network-based parallel filesystem. Based on this analysis, we then present LMDBIO, our optimized I/O plugin for Caffe that takes into account the data access pattern of Caffe in order to vastly improve I/O performance. Our experimental results show that LMDBIO can improve the overall execution time of Caffe by nearly 20-fold in some cases.« less

Observing the Birth and evolution of Galaxies - the ATCA-AKARI-ASTE/AzTEC deep South Ecliptic Pole Field

NASA Astrophysics Data System (ADS)

White, Glenn; Kohno, Kotaro; Matsuhara, Hideo; Matsuura, Shuji; Hanami, Hitoshi; Lee, Hyung Mok; Pearson, Chris; Takagi, Toshi; Serjeant, Stephen; Jeong, Woongseob; Oyabu, Shinki; Shirahata, Mai; Nakanishi, Kouichiro; Figueredo, Elysandra; Etxaluze, Mireya

2007-04-01

We propose deep 20 cm observations supporting the AKARI (3-160 micron)/ASTE/AzTEC (1.1 mm) SEP ultra deep ('Oyabu Field') survey of an extremely low cirrus region at the South Ecliptic Pole. Our combined IR/mm/Radio survey addresses the questions: How do protogalaxies and protospheroids form and evolve? How do AGN link with ULIRGs in their birth and evolution? What is the nature of the mm/submm extragalactic source population? We will address these by sampling the star formation history in the early universe to at least z~2. Compared to other Deep Surveys, a) AKARI multi-band IR measurements allow precision photo-z estimates of optically obscured objects, b) our multi-waveband contiguous area will mitigate effects of cosmic variance, c) the low cirrus noise at the SEP (< 0.08 MJy/sr) rivals that of the Lockman Hole "Astronomy's other ultra-deep 'cosmological window'", and d) our coverage of four FIR bands will characterise the far-IR dust emission hump of our starburst galaxies better than SPITZER's two MIPS bands allow. The ATCA data are crucial to galaxy identification, and determining the star formation rates and intrinsic luminosities through this unique Southern cosmological window.

76 FR 74757 - Fisheries of the Caribbean, Gulf of Mexico, and South Atlantic; Comprehensive Annual Catch Limit...

Federal Register 2010, 2011, 2012, 2013, 2014

2011-12-01

... place selected snapper-grouper species into the complexes for: Deep-water species (yellowedge grouper... snapper); shallow-water groupers (red hind, rock hind, yellowmouth grouper, yellowfin grouper, coney, and... ``South Atlantic shallow- water grouper (SASWG)'' is revised to read as follows: Sec. 622.2 Definitions...

Prediction of Bispectral Index during Target-controlled Infusion of Propofol and Remifentanil: A Deep Learning Approach.

PubMed

Lee, Hyung-Chul; Ryu, Ho-Geol; Chung, Eun-Jin; Jung, Chul-Woo

2018-03-01

The discrepancy between predicted effect-site concentration and measured bispectral index is problematic during intravenous anesthesia with target-controlled infusion of propofol and remifentanil. We hypothesized that bispectral index during total intravenous anesthesia would be more accurately predicted by a deep learning approach. Long short-term memory and the feed-forward neural network were sequenced to simulate the pharmacokinetic and pharmacodynamic parts of an empirical model, respectively, to predict intraoperative bispectral index during combined use of propofol and remifentanil. Inputs of long short-term memory were infusion histories of propofol and remifentanil, which were retrieved from target-controlled infusion pumps for 1,800 s at 10-s intervals. Inputs of the feed-forward network were the outputs of long short-term memory and demographic data such as age, sex, weight, and height. The final output of the feed-forward network was the bispectral index. The performance of bispectral index prediction was compared between the deep learning model and previously reported response surface model. The model hyperparameters comprised 8 memory cells in the long short-term memory layer and 16 nodes in the hidden layer of the feed-forward network. The model training and testing were performed with separate data sets of 131 and 100 cases. The concordance correlation coefficient (95% CI) were 0.561 (0.560 to 0.562) in the deep learning model, which was significantly larger than that in the response surface model (0.265 [0.263 to 0.266], P < 0.001). The deep learning model-predicted bispectral index during target-controlled infusion of propofol and remifentanil more accurately compared to the traditional model. The deep learning approach in anesthetic pharmacology seems promising because of its excellent performance and extensibility.

A Multiobjective Sparse Feature Learning Model for Deep Neural Networks.

PubMed

Gong, Maoguo; Liu, Jia; Li, Hao; Cai, Qing; Su, Linzhi

2015-12-01

Hierarchical deep neural networks are currently popular learning models for imitating the hierarchical architecture of human brain. Single-layer feature extractors are the bricks to build deep networks. Sparse feature learning models are popular models that can learn useful representations. But most of those models need a user-defined constant to control the sparsity of representations. In this paper, we propose a multiobjective sparse feature learning model based on the autoencoder. The parameters of the model are learnt by optimizing two objectives, reconstruction error and the sparsity of hidden units simultaneously to find a reasonable compromise between them automatically. We design a multiobjective induced learning procedure for this model based on a multiobjective evolutionary algorithm. In the experiments, we demonstrate that the learning procedure is effective, and the proposed multiobjective model can learn useful sparse features.

Deep Convolutional Neural Network-Based Early Automated Detection of Diabetic Retinopathy Using Fundus Image.

PubMed

Xu, Kele; Feng, Dawei; Mi, Haibo

2017-11-23

The automatic detection of diabetic retinopathy is of vital importance, as it is the main cause of irreversible vision loss in the working-age population in the developed world. The early detection of diabetic retinopathy occurrence can be very helpful for clinical treatment; although several different feature extraction approaches have been proposed, the classification task for retinal images is still tedious even for those trained clinicians. Recently, deep convolutional neural networks have manifested superior performance in image classification compared to previous handcrafted feature-based image classification methods. Thus, in this paper, we explored the use of deep convolutional neural network methodology for the automatic classification of diabetic retinopathy using color fundus image, and obtained an accuracy of 94.5% on our dataset, outperforming the results obtained by using classical approaches.

Future Mission Trends and their Implications for the Deep Space Network

NASA Technical Reports Server (NTRS)

Abraham, Douglas S.

2006-01-01

Planning for the upgrade and/or replacement of Deep Space Network (DSN) assets that typically operate for forty or more years necessitates understanding potential customer needs as far into the future as possible. This paper describes the methodology Deep Space Network (DSN) planners use to develop this understanding, some key future mission trends that have emerged from application of this methodology, and the implications of the trends for the DSN's future evolution. For NASA's current plans out to 2030, these trends suggest the need to accommodate: three times as many communication links, downlink rates two orders of magnitude greater than today's, uplink rates some four orders of magnitude greater, and end-to-end link difficulties two-to-three orders of magnitude greater. To meet these challenges, both DSN capacity and capability will need to increase.

32 GHz Celestial Reference Frame Survey for Dec < -45 deg.

NASA Astrophysics Data System (ADS)

Horiuchi, Shinji; Phillips, Chris; Stevens, Jamie; Jacobs, Christopher; Sotuela, Ioana; Garcia miro, Cristina

2014-04-01

(We resubmit this proposal to extend from the previous semester. The 24 hour blocks for ATCA and Mopra were granted in May 2012 but canceled because fringe test before the scheduled experiment failed although fringes were detected between Mopra and Tidbinbilla. During the last scheduled LBA session for this project we discovered ATCA/Mopra had an issue with frequency standard, which has now been resolved.) We propose to conduct a LBA survey of compact radio sources at 32 GHz near the south pole region. This is the first attempt to fill the gap in the existing 32 GHz catalogue establish by NASA Deep Space Network toward completing the full sky celestial reference frame at 32 GHz. The catalogue will be used for future spacecraft navigation by NASA and other space agencies as well as for radio astronomical observations with southern radio telescope arrays such as ATCA and LBA.

Single image super-resolution based on convolutional neural networks

NASA Astrophysics Data System (ADS)

Zou, Lamei; Luo, Ming; Yang, Weidong; Li, Peng; Jin, Liujia

2018-03-01

We present a deep learning method for single image super-resolution (SISR). The proposed approach learns end-to-end mapping between low-resolution (LR) images and high-resolution (HR) images. The mapping is represented as a deep convolutional neural network which inputs the LR image and outputs the HR image. Our network uses 5 convolution layers, which kernels size include 5×5, 3×3 and 1×1. In our proposed network, we use residual-learning and combine different sizes of convolution kernels at the same layer. The experiment results show that our proposed method performs better than the existing methods in reconstructing quality index and human visual effects on benchmarked images.

DEEP UNDERGROUND NEUTRINO EXPERIMENT

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

Wilson, Robert J.

2016-03-03

The Deep Underground Neutrino Experiment (DUNE) collaboration will perform an experiment centered on accelerator-based long-baseline neutrino studies along with nucleon decay and topics in neutrino astrophysics. It will consist of a modular 40-kt (fiducial) mass liquid argon TPC detector located deep underground at the Sanford Underground Research Facility in South Dakota and a high-resolution near detector at Fermilab in Illinois. This conguration provides a 1300-km baseline in a megawatt-scale neutrino beam provided by the Fermilab- hosted international Long-Baseline Neutrino Facility.

Seismicity, shear failure and modes of deformation in deep subduction zones

NASA Technical Reports Server (NTRS)

Lundgren, Paul R.; Giardini, Domenico

1992-01-01

The joint hypocentral determination method is used to relocate deep seismicity reported in the International Seismological Center catalog for earthquakes deeper than 400 km in the Honshu, Bonin, Mariannas, Java, Banda, and South America subduction zones. Each deep seismic zone is found to display planar features of seismicity parallel to the Harvard centroid-moment tensor nodal planes, which are identified as planes of shear failure. The sense of displacement on these planes is one of resistance to deeper penetration.

Corrigendum to ;Stirring by deep cyclones and the evolution of Denmark strait overflow water observed at Line W; [Deep-Sea Res. I 109, 10-26

NASA Astrophysics Data System (ADS)

Andres, M.; Toole, J. M.; Torres, D. J.; Smethie, W. M.; Joyce, T. M.; Curry, R. G.

2017-03-01

The Line W program was a 10-year study (2004-2014) to investigate variability in the Deep Western Boundary Current (DWBC) and the nearby ocean interior south of New England. Line W stretches from the Middle Atlantic Bight continental slope southeastward towards Bermuda along a satellite altimeter track and is roughly orthogonal to the 2500-3500 m isobaths along the continental slope here (Fig. 1a).

Using statistical and artificial neural network models to forecast potentiometric levels at a deep well in South Texas

NASA Astrophysics Data System (ADS)

Uddameri, V.

2007-01-01

Reliable forecasts of monthly and quarterly fluctuations in groundwater levels are necessary for short- and medium-term planning and management of aquifers to ensure proper service of seasonal demands within a region. Development of physically based transient mathematical models at this time scale poses considerable challenges due to lack of suitable data and other uncertainties. Artificial neural networks (ANN) possess flexible mathematical structures and are capable of mapping highly nonlinear relationships. Feed-forward neural network models were constructed and trained using the back-percolation algorithm to forecast monthly and quarterly time-series water levels at a well that taps into the deeper Evangeline formation of the Gulf Coast aquifer in Victoria, TX. Unlike unconfined formations, no causal relationships exist between water levels and hydro-meteorological variables measured near the vicinity of the well. As such, an endogenous forecasting model using dummy variables to capture short-term seasonal fluctuations and longer-term (decadal) trends was constructed. The root mean square error, mean absolute deviation and correlation coefficient ( R) were noted to be 1.40, 0.33 and 0.77 m, respectively, for an evaluation dataset of quarterly measurements and 1.17, 0.46, and 0.88 m for an evaluative monthly dataset not used to train or test the model. These statistics were better for the ANN model than those developed using statistical regression techniques.

77 FR 15915 - Fisheries of the Caribbean, Gulf of Mexico, and South Atlantic; Comprehensive Annual Catch Limit...

Federal Register 2010, 2011, 2012, 2013, 2014

2012-03-16

...-grouper species into the complexes for: Deep-water species (yellowedge grouper, blueline tilefish, silk snapper, misty grouper, sand tilefish, queen snapper, black snapper, and blackfin snapper); shallow-water... caught in very deep water (1,476-1,969 ft (450- 600 m)), it is assumed that all incidentally caught...

Building America Case Study: Pilot Demonstration of Phased Energy Efficiency Retrofits: Deep Retrofits, Central and South Florida

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

D. Parker, K. Sutherland, D. Chasar, J. Montemurno, B. Amos, J. Kono

2017-02-01

The Florida Solar Energy Center (FSEC), in collaboration with Florida Power & Light (FPL), is pursuing a phased residential energy-efficiency retrofit program in Florida. Researchers are looking to establish the impacts of technologies of two retrofit packages -- shallow and deep -- on annual energy and peak energy reductions.

Building America Case Study: Pilot Demonstration of Phased Energy Efficiency Retrofits: Deep Retrofits, Central and South Florida

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

2017-02-22

The Florida Solar Energy Center (FSEC), in collaboration with Florida Power & Light (FPL), is pursuing a phased residential energy-efficiency retrofit program in Florida. Researchers are looking to establish the impacts of technologies of two retrofit packages -- shallow and deep -- on annual energy and peak energy reductions.

Conceptual Tutoring Software for Promoting Deep Learning: A Case Study

ERIC Educational Resources Information Center

Stott, Angela; Hattingh, Annemarie

2015-01-01

The paper presents a case study of the use of conceptual tutoring software to promote deep learning of the scientific concept of density among 50 final year pre-service student teachers in a natural sciences course in a South African university. Individually-paced electronic tutoring is potentially an effective way of meeting the students' varied…

Deep space network software cost estimation model

NASA Technical Reports Server (NTRS)

Tausworthe, R. C.

1981-01-01

A parametric software cost estimation model prepared for Jet PRopulsion Laboratory (JPL) Deep Space Network (DSN) Data System implementation tasks is described. The resource estimation mdel modifies and combines a number of existing models. The model calibrates the task magnitude and difficulty, development environment, and software technology effects through prompted responses to a set of approximately 50 questions. Parameters in the model are adjusted to fit JPL software life-cycle statistics.

LEONA: Transient Luminous Event and Thunderstorm High Energy Emission Collaborative Network in Latin America

NASA Astrophysics Data System (ADS)

Sao Sabbas, F. T.

2012-12-01

This project has the goal of establishing the Collaborative Network LEONA, to study the electrodynamical coupling of the atmospheric layers signaled by Transient Luminous Events - TLEs and high energy emissions from thunderstorms. We will develop and install a remotely controlled network of cameras to perform TLE observations in different locations in South America and one neutron detector in southern Brazil. The camera network will allow building a continuous data set of the phenomena studied in this continent. The first two trial units of the camera network are already installed, in Brazil and Peru, and two more will be installed until December 2012, in Argentina and Brazil. We expect to determine the TLE geographic distribution, occurrence rate, morphology, and possible coupling with other geophysical phenomena in South America, such as the South Atlantic Magnetic Anomaly - SAMA. We also expect to study thunderstorm neutron emissions in a region of intense electrical activity, measuring neutron fluxes with high time resolution simultaneously with TLEs and lightning for the first time in South America. Using an intensified high-speed camera for TLE observation during 2 campaigns we expect to be able to determine the duration and spatial- temporal development of the TLEs observed, to study the structure and initiation of sprites and to measure the velocity of development of sprite structures and the sprite delay. The camera was acquired via the FAPESP project DEELUMINOS (2005-2010), which also nucleated our research group Atmospheric Electrodynamical Coupling - ACATMOS. LEONA will nucleate this research in other institutions in Brazil and other countries in South America, providing continuity for this important research in our region. The camera network will be an unique tool to perform consistent long term TLE observation, and in fact is the only way to accumulate a data set for a climatological study of South America, since satellite instrumentation turns off in this region to avoid damages due to the South Atlantic Magnetic Anomaly - SAMA. Thus this project is not only a potential benchmark in TLE research by creating a collaborative network in Latin America and nucleating this research locally, it is also strategic since LEONA's camera network will be able to provide extremely valuable information to fill up this gap that most satellite measurements have.

EMG-Based Estimation of Limb Movement Using Deep Learning With Recurrent Convolutional Neural Networks.

PubMed

Xia, Peng; Hu, Jie; Peng, Yinghong

2017-10-25

A novel model based on deep learning is proposed to estimate kinematic information for myoelectric control from multi-channel electromyogram (EMG) signals. The neural information of limb movement is embedded in EMG signals that are influenced by all kinds of factors. In order to overcome the negative effects of variability in signals, the proposed model employs the deep architecture combining convolutional neural networks (CNNs) and recurrent neural networks (RNNs). The EMG signals are transformed to time-frequency frames as the input to the model. The limb movement is estimated by the model that is trained with the gradient descent and backpropagation procedure. We tested the model for simultaneous and proportional estimation of limb movement in eight healthy subjects and compared it with support vector regression (SVR) and CNNs on the same data set. The experimental studies show that the proposed model has higher estimation accuracy and better robustness with respect to time. The combination of CNNs and RNNs can improve the model performance compared with using CNNs alone. The model of deep architecture is promising in EMG decoding and optimization of network structures can increase the accuracy and robustness. © 2017 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.

Defining the natural fracture network in a shale gas play and its cover succession: The case of the Utica Shale in eastern Canada

NASA Astrophysics Data System (ADS)

Ladevèze, P.; Séjourné, S.; Rivard, C.; Lavoie, D.; Lefebvre, R.; Rouleau, A.

2018-03-01

In the St. Lawrence sedimentary platform (eastern Canada), very little data are available between shallow fresh water aquifers and deep geological hydrocarbon reservoir units (here referred to as the intermediate zone). Characterization of this intermediate zone is crucial, as the latter controls aquifer vulnerability to operations carried out at depth. In this paper, the natural fracture networks in shallow aquifers and in the Utica shale gas reservoir are documented in an attempt to indirectly characterize the intermediate zone. This study used structural data from outcrops, shallow observation well logs and deep shale gas well logs to propose a conceptual model of the natural fracture network. Shallow and deep fractures were categorized into three sets of steeply-dipping fractures and into a set of bedding-parallel fractures. Some lithological and structural controls on fracture distribution were identified. The regional geologic history and similarities between the shallow and deep fracture datasets allowed the extrapolation of the fracture network characterization to the intermediate zone. This study thus highlights the benefits of using both datasets simultaneously, while they are generally interpreted separately. Recommendations are also proposed for future environmental assessment studies in which the existence of preferential flow pathways and potential upward fluid migration toward shallow aquifers need to be identified.

Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition

PubMed Central

Ordóñez, Francisco Javier; Roggen, Daniel

2016-01-01

Human activity recognition (HAR) tasks have traditionally been solved using engineered features obtained by heuristic processes. Current research suggests that deep convolutional neural networks are suited to automate feature extraction from raw sensor inputs. However, human activities are made of complex sequences of motor movements, and capturing this temporal dynamics is fundamental for successful HAR. Based on the recent success of recurrent neural networks for time series domains, we propose a generic deep framework for activity recognition based on convolutional and LSTM recurrent units, which: (i) is suitable for multimodal wearable sensors; (ii) can perform sensor fusion naturally; (iii) does not require expert knowledge in designing features; and (iv) explicitly models the temporal dynamics of feature activations. We evaluate our framework on two datasets, one of which has been used in a public activity recognition challenge. Our results show that our framework outperforms competing deep non-recurrent networks on the challenge dataset by 4% on average; outperforming some of the previous reported results by up to 9%. Our results show that the framework can be applied to homogeneous sensor modalities, but can also fuse multimodal sensors to improve performance. We characterise key architectural hyperparameters’ influence on performance to provide insights about their optimisation. PMID:26797612

Simulation of noisy dynamical system by Deep Learning

NASA Astrophysics Data System (ADS)

Yeo, Kyongmin

2017-11-01

Deep learning has attracted huge attention due to its powerful representation capability. However, most of the studies on deep learning have been focused on visual analytics or language modeling and the capability of the deep learning in modeling dynamical systems is not well understood. In this study, we use a recurrent neural network to model noisy nonlinear dynamical systems. In particular, we use a long short-term memory (LSTM) network, which constructs internal nonlinear dynamics systems. We propose a cross-entropy loss with spatial ridge regularization to learn a non-stationary conditional probability distribution from a noisy nonlinear dynamical system. A Monte Carlo procedure to perform time-marching simulations by using the LSTM is presented. The behavior of the LSTM is studied by using noisy, forced Van der Pol oscillator and Ikeda equation.

Learning Traffic as Images: A Deep Convolutional Neural Network for Large-Scale Transportation Network Speed Prediction.

PubMed

Ma, Xiaolei; Dai, Zhuang; He, Zhengbing; Ma, Jihui; Wang, Yong; Wang, Yunpeng

2017-04-10

This paper proposes a convolutional neural network (CNN)-based method that learns traffic as images and predicts large-scale, network-wide traffic speed with a high accuracy. Spatiotemporal traffic dynamics are converted to images describing the time and space relations of traffic flow via a two-dimensional time-space matrix. A CNN is applied to the image following two consecutive steps: abstract traffic feature extraction and network-wide traffic speed prediction. The effectiveness of the proposed method is evaluated by taking two real-world transportation networks, the second ring road and north-east transportation network in Beijing, as examples, and comparing the method with four prevailing algorithms, namely, ordinary least squares, k-nearest neighbors, artificial neural network, and random forest, and three deep learning architectures, namely, stacked autoencoder, recurrent neural network, and long-short-term memory network. The results show that the proposed method outperforms other algorithms by an average accuracy improvement of 42.91% within an acceptable execution time. The CNN can train the model in a reasonable time and, thus, is suitable for large-scale transportation networks.

Learning Traffic as Images: A Deep Convolutional Neural Network for Large-Scale Transportation Network Speed Prediction

PubMed Central

Ma, Xiaolei; Dai, Zhuang; He, Zhengbing; Ma, Jihui; Wang, Yong; Wang, Yunpeng

2017-01-01

This paper proposes a convolutional neural network (CNN)-based method that learns traffic as images and predicts large-scale, network-wide traffic speed with a high accuracy. Spatiotemporal traffic dynamics are converted to images describing the time and space relations of traffic flow via a two-dimensional time-space matrix. A CNN is applied to the image following two consecutive steps: abstract traffic feature extraction and network-wide traffic speed prediction. The effectiveness of the proposed method is evaluated by taking two real-world transportation networks, the second ring road and north-east transportation network in Beijing, as examples, and comparing the method with four prevailing algorithms, namely, ordinary least squares, k-nearest neighbors, artificial neural network, and random forest, and three deep learning architectures, namely, stacked autoencoder, recurrent neural network, and long-short-term memory network. The results show that the proposed method outperforms other algorithms by an average accuracy improvement of 42.91% within an acceptable execution time. The CNN can train the model in a reasonable time and, thus, is suitable for large-scale transportation networks. PMID:28394270

HIV-1 subtype F1 epidemiological networks among Italian heterosexual males are associated with introduction events from South America.

PubMed

Lai, Alessia; Simonetti, Francesco R; Zehender, Gianguglielmo; De Luca, Andrea; Micheli, Valeria; Meraviglia, Paola; Corsi, Paola; Bagnarelli, Patrizia; Almi, Paolo; Zoncada, Alessia; Paolucci, Stefania; Gonnelli, Angela; Colao, Grazia; Tacconi, Danilo; Franzetti, Marco; Ciccozzi, Massimo; Zazzi, Maurizio; Balotta, Claudia

2012-01-01

About 40% of the Italian HIV-1 epidemic due to non-B variants is sustained by F1 clade, which circulates at high prevalence in South America and Eastern Europe. Aim of this study was to define clade F1 origin, population dynamics and epidemiological networks through phylogenetic approaches. We analyzed pol sequences of 343 patients carrying F1 subtype stored in the ARCA database from 1998 to 2009. Citizenship of patients was as follows: 72.6% Italians, 9.3% South Americans and 7.3% Rumanians. Heterosexuals, Homo-bisexuals, Intravenous Drug Users accounted for 58.1%, 24.0% and 8.8% of patients, respectively. Phylogenetic analysis indicated that 70% of sequences clustered in 27 transmission networks. Two distinct groups were identified; the first clade, encompassing 56 sequences, included all Rumanian patients. The second group involved the remaining clusters and included 10 South American Homo-bisexuals in 9 distinct clusters. Heterosexual modality of infection was significantly associated with the probability to be detected in transmission networks. Heterosexuals were prevalent either among Italians (67.2%) or Rumanians (50%); by contrast, Homo-bisexuals accounted for 71.4% of South Americans. Among patients with resistant strains the proportion of clustering sequences was 57.1%, involving 14 clusters (51.8%). Resistance in clusters tended to be higher in South Americans (28.6%) compared to Italian (17.7%) and Rumanian patients (14.3%). A striking proportion of epidemiological networks could be identified in heterosexuals carrying F1 subtype residing in Italy. Italian Heterosexual males predominated within epidemiological clusters while foreign patients were mainly Heterosexual Rumanians, both males and females, and South American Homo-bisexuals. Tree topology suggested that F1 variant from South America gave rise to the Italian F1 epidemic through multiple introduction events. The contact tracing also revealed an unexpected burden of resistance in epidemiological clusters underlying the need of public interventions to limit the spread of non-B subtypes and transmitted drug resistance.

Rapid deformation of the South flank of kilauea volcano, hawaii.

PubMed

Owen, S; Segall, P; Freymueller, J; Mikijus, A; Denlinger, R; Arnadóttir, T; Sako, M; Bürgmann, R

1995-03-03

The south flank of Kilauea volcano has experienced two large [magnitude (M) 7.2 and M 6.1] earthquakes in the past two decades. Global Positioning System measurements conducted between 1990 and 1993 reveal seaward displacements of Kilauea's central south flank at rates of up to about 10 centimeters per year. In contrast, the northern side of the volcano and the distal ends of the south flank did not displace significantly. The observations can be explained by slip on a low-angle fault beneath the south flank combined with dilation deep within Kilauea's rift system, both at rates of at least 15 centimeters per year.

ICADx: interpretable computer aided diagnosis of breast masses

NASA Astrophysics Data System (ADS)

Kim, Seong Tae; Lee, Hakmin; Kim, Hak Gu; Ro, Yong Man

2018-02-01

In this study, a novel computer aided diagnosis (CADx) framework is devised to investigate interpretability for classifying breast masses. Recently, a deep learning technology has been successfully applied to medical image analysis including CADx. Existing deep learning based CADx approaches, however, have a limitation in explaining the diagnostic decision. In real clinical practice, clinical decisions could be made with reasonable explanation. So current deep learning approaches in CADx are limited in real world deployment. In this paper, we investigate interpretability in CADx with the proposed interpretable CADx (ICADx) framework. The proposed framework is devised with a generative adversarial network, which consists of interpretable diagnosis network and synthetic lesion generative network to learn the relationship between malignancy and a standardized description (BI-RADS). The lesion generative network and the interpretable diagnosis network compete in an adversarial learning so that the two networks are improved. The effectiveness of the proposed method was validated on public mammogram database. Experimental results showed that the proposed ICADx framework could provide the interpretability of mass as well as mass classification. It was mainly attributed to the fact that the proposed method was effectively trained to find the relationship between malignancy and interpretations via the adversarial learning. These results imply that the proposed ICADx framework could be a promising approach to develop the CADx system.

Deep Neural Networks Based Recognition of Plant Diseases by Leaf Image Classification.

PubMed

Sladojevic, Srdjan; Arsenovic, Marko; Anderla, Andras; Culibrk, Dubravko; Stefanovic, Darko

2016-01-01

The latest generation of convolutional neural networks (CNNs) has achieved impressive results in the field of image classification. This paper is concerned with a new approach to the development of plant disease recognition model, based on leaf image classification, by the use of deep convolutional networks. Novel way of training and the methodology used facilitate a quick and easy system implementation in practice. The developed model is able to recognize 13 different types of plant diseases out of healthy leaves, with the ability to distinguish plant leaves from their surroundings. According to our knowledge, this method for plant disease recognition has been proposed for the first time. All essential steps required for implementing this disease recognition model are fully described throughout the paper, starting from gathering images in order to create a database, assessed by agricultural experts. Caffe, a deep learning framework developed by Berkley Vision and Learning Centre, was used to perform the deep CNN training. The experimental results on the developed model achieved precision between 91% and 98%, for separate class tests, on average 96.3%.

Deep Neural Networks Based Recognition of Plant Diseases by Leaf Image Classification

PubMed Central

Sladojevic, Srdjan; Arsenovic, Marko; Culibrk, Dubravko; Stefanovic, Darko

2016-01-01

The latest generation of convolutional neural networks (CNNs) has achieved impressive results in the field of image classification. This paper is concerned with a new approach to the development of plant disease recognition model, based on leaf image classification, by the use of deep convolutional networks. Novel way of training and the methodology used facilitate a quick and easy system implementation in practice. The developed model is able to recognize 13 different types of plant diseases out of healthy leaves, with the ability to distinguish plant leaves from their surroundings. According to our knowledge, this method for plant disease recognition has been proposed for the first time. All essential steps required for implementing this disease recognition model are fully described throughout the paper, starting from gathering images in order to create a database, assessed by agricultural experts. Caffe, a deep learning framework developed by Berkley Vision and Learning Centre, was used to perform the deep CNN training. The experimental results on the developed model achieved precision between 91% and 98%, for separate class tests, on average 96.3%. PMID:27418923

Deep greedy learning under thermal variability in full diurnal cycles

NASA Astrophysics Data System (ADS)

Rauss, Patrick; Rosario, Dalton

2017-08-01

We study the generalization and scalability behavior of a deep belief network (DBN) applied to a challenging long-wave infrared hyperspectral dataset, consisting of radiance from several manmade and natural materials within a fixed site located 500 m from an observation tower. The collections cover multiple full diurnal cycles and include different atmospheric conditions. Using complementary priors, a DBN uses a greedy algorithm that can learn deep, directed belief networks one layer at a time and has two layers form to provide undirected associative memory. The greedy algorithm initializes a slower learning procedure, which fine-tunes the weights, using a contrastive version of the wake-sleep algorithm. After fine-tuning, a network with three hidden layers forms a very good generative model of the joint distribution of spectral data and their labels, despite significant data variability between and within classes due to environmental and temperature variation occurring within and between full diurnal cycles. We argue, however, that more questions than answers are raised regarding the generalization capacity of these deep nets through experiments aimed at investigating their training and augmented learning behavior.

Evaluating the Generalization Value of Process-based Models in a Deep-in-time Machine Learning framework

NASA Astrophysics Data System (ADS)

Shen, C.; Fang, K.

2017-12-01

Deep Learning (DL) methods have made revolutionary strides in recent years. A core value proposition of DL is that abstract notions and patterns can be extracted purely from data, without the need for domain expertise. Process-based models (PBM), on the other hand, can be regarded as repositories of human knowledge or hypotheses about how systems function. Here, through computational examples, we argue that there is merit in integrating PBMs with DL due to the imbalance and lack of data in many situations, especially in hydrology. We trained a deep-in-time neural network, the Long Short-Term Memory (LSTM), to learn soil moisture dynamics from Soil Moisture Active Passive (SMAP) Level 3 product. We show that when PBM solutions are integrated into LSTM, the network is able to better generalize across regions. LSTM is able to better utilize PBM solutions than simpler statistical methods. Our results suggest PBMs have generalization value which should be carefully assessed and utilized. We also emphasize that when properly regularized, the deep network is robust and is of superior testing performance compared to simpler methods.

The Importance of the Effective Utilization of Women at Arms

DTIC Science & Technology

2015-03-01

Network Presented in Zagreb ,” South Eastern and Eastern Europe Clearinghouse for the Control of Small Arms and Light Weapons, November 5, 2010, http...for the Control of Small Arms and Light Weapons. “Report Establishing the South East Europe Women Police Officers Network Presented in Zagreb

Late Departures from Paper-Based to Supported Networked Learning in South Africa: Lessons Learned

ERIC Educational Resources Information Center

Kok, Illasha; Beter, Petra; Esterhuizen, Hennie

2018-01-01

Fragmented connectivity in South Africa is the dominant barrier for digitising initiatives. New insights surfaced when a university-based nursing programme introduced tablets within a supportive network learning environment. A qualitative, explorative design investigated adult nurses' experiences of the realities when moving from paper-based…

Silicon Isotope Variations in Giant Spicules of the Deep-sea Sponge Monorhaphis chuni

NASA Astrophysics Data System (ADS)

Jochum, K. P.; Schuessler, J. A.; Wang, X.; Müller, W. E.; Andreae, M. O.

2012-12-01

The astonishing longevity of the deep-sea sponge Monorhaphis chuni and the stability of their spicules (Wang et al. 2009) provide the potential that single giant basal spicules can be used as paleoenvironmental archives spanning the entire Holocene (Jochum et al. 2012). According to Wille et al. (2010), the Si isotope fractionation is influenced by seawater Si concentration with lower values associated with sponges collected from waters high in Si. In order to track possible secular variations during the last 10000 years in the deep sea, we have therefore determined Si isotope ratios and trace element ratios along center-to-surface sections at a high resolution by femtosecond LA-(MC)-ICP-MS. Samples came from different locations of the East and South China Sea as well as the South Pacific Ocean (near New Caledonia) and were collected at depths between 1100 m and 2100 m. The external reproducibility of the fs LA-(MC)-ICP-MS method was found to be 0.14 ‰ and 0.27 ‰ (2 SD) for δ29Si and δ30Si, respectively. The relative uncertainty on trace element abundance ratios, such as Mg/Ca, is about 5 % (RSD). Significant variations in Si isotope ratios were observed in the giant spicules Q-B and SCS-4 from the East and South China Sea, respectively. The δ30Si values for the largest spicule collected so far (SCS-4, 2.5 m long) from a depth of 2100 m in the South China Sea, span a large range from -1.9 to -3.7 ‰. No obvious trend in Si isotope variability outside external reproducibility could be identified in smaller and presumably younger spicules; average δ30Si values of 4 different segments of the spicule MC from the South China Sea are about -1.3 ‰. Low δ30Si values of about -0.88 ‰ are found in the giant spicule V from the South Pacific. Mg/Ca ratios of most spicules show small, but significant trends from higher values at the rim to lower values in the core, which can be interpreted as an increase in seawater temperature of several degrees Celsius during the lifespan of the sponges. The different Si isotopic compositions in the deep sea may be caused by regional and vertical differences of dissolved Si controlled by biological productivity and ocean circulation. Submarine weathering at sea floor hydrothermal areas provides additional silicic acid in the deep reservoir. Jochum et al. (2012), Chem. Geol. 300-301, 143-151 Wang et al. (2009), Int. Rev. Cell. Mol. Biol. 273, 69-115 Wille et al. (2010), Earth Planet. Sci. Lett. 292, 281-289

The Impact of the Great Migration on Mortality of African Americans: Evidence from the Deep South

PubMed Central

Black, Dan A.; Sanders, Seth G.; Taylor, Evan J.

2015-01-01

The Great Migration—the massive migration of African Americans out of the rural South to largely urban locations in the North, Midwest, and West—was a landmark event in U.S. history. Our paper shows that this migration increased mortality of African Americans born in the early twentieth century South. This inference comes from an analysis that uses proximity of birthplace to railroad lines as an instrument for migration. PMID:26345146

Shallow Transits—Deep Learning. I. Feasibility Study of Deep Learning to Detect Periodic Transits of Exoplanets

NASA Astrophysics Data System (ADS)

Zucker, Shay; Giryes, Raja

2018-04-01

Transits of habitable planets around solar-like stars are expected to be shallow, and to have long periods, which means low information content. The current bottleneck in the detection of such transits is caused in large part by the presence of red (correlated) noise in the light curves obtained from the dedicated space telescopes. Based on the groundbreaking results deep learning achieves in many signal and image processing applications, we propose to use deep neural networks to solve this problem. We present a feasibility study, in which we applied a convolutional neural network on a simulated training set. The training set comprised light curves received from a hypothetical high-cadence space-based telescope. We simulated the red noise by using Gaussian Processes with a wide variety of hyper-parameters. We then tested the network on a completely different test set simulated in the same way. Our study proves that very difficult cases can indeed be detected. Furthermore, we show how detection trends can be studied and detection biases quantified. We have also checked the robustness of the neural-network performance against practical artifacts such as outliers and discontinuities, which are known to affect space-based high-cadence light curves. Future work will allow us to use the neural networks to characterize the transit model and identify individual transits. This new approach will certainly be an indispensable tool for the detection of habitable planets in the future planet-detection space missions such as PLATO.

Recurrent neural networks for breast lesion classification based on DCE-MRIs

NASA Astrophysics Data System (ADS)

Antropova, Natasha; Huynh, Benjamin; Giger, Maryellen

2018-02-01

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) plays a significant role in breast cancer screening, cancer staging, and monitoring response to therapy. Recently, deep learning methods are being rapidly incorporated in image-based breast cancer diagnosis and prognosis. However, most of the current deep learning methods make clinical decisions based on 2-dimentional (2D) or 3D images and are not well suited for temporal image data. In this study, we develop a deep learning methodology that enables integration of clinically valuable temporal components of DCE-MRIs into deep learning-based lesion classification. Our work is performed on a database of 703 DCE-MRI cases for the task of distinguishing benign and malignant lesions, and uses the area under the ROC curve (AUC) as the performance metric in conducting that task. We train a recurrent neural network, specifically a long short-term memory network (LSTM), on sequences of image features extracted from the dynamic MRI sequences. These features are extracted with VGGNet, a convolutional neural network pre-trained on a large dataset of natural images ImageNet. The features are obtained from various levels of the network, to capture low-, mid-, and high-level information about the lesion. Compared to a classification method that takes as input only images at a single time-point (yielding an AUC = 0.81 (se = 0.04)), our LSTM method improves lesion classification with an AUC of 0.85 (se = 0.03).

Computational ghost imaging using deep learning

NASA Astrophysics Data System (ADS)

Shimobaba, Tomoyoshi; Endo, Yutaka; Nishitsuji, Takashi; Takahashi, Takayuki; Nagahama, Yuki; Hasegawa, Satoki; Sano, Marie; Hirayama, Ryuji; Kakue, Takashi; Shiraki, Atsushi; Ito, Tomoyoshi

2018-04-01

Computational ghost imaging (CGI) is a single-pixel imaging technique that exploits the correlation between known random patterns and the measured intensity of light transmitted (or reflected) by an object. Although CGI can obtain two- or three-dimensional images with a single or a few bucket detectors, the quality of the reconstructed images is reduced by noise due to the reconstruction of images from random patterns. In this study, we improve the quality of CGI images using deep learning. A deep neural network is used to automatically learn the features of noise-contaminated CGI images. After training, the network is able to predict low-noise images from new noise-contaminated CGI images.

NASA deep space network operations planning and preparation

NASA Technical Reports Server (NTRS)

Jensen, W. N.

1982-01-01

The responsibilities and structural organization of the Operations Planning Group of NASA Deep Space Network (DSN) Operations are outlined. The Operations Planning group establishes an early interface with a user's planning organization to educate the user on DSN capabilities and limitations for deep space tracking support. A team of one or two individuals works through all phases of the spacecraft launch and also provides planning and preparation for specific events such as planetary encounters. Coordinating interface is also provided for nonflight projects such as radio astronomy and VLBI experiments. The group is divided into a Long Range Support Planning element and a Near Term Operations Coordination element.

The Deep Space Network as an instrument for radio science research

NASA Technical Reports Server (NTRS)

Asmar, S. W.; Renzetti, N. A.

1993-01-01

Radio science experiments use radio links between spacecraft and sensor instrumentation that is implemented in the Deep Space Network. The deep space communication complexes along with the telecommunications subsystem on board the spacecraft constitute the major elements of the radio science instrumentation. Investigators examine small changes in the phase and/or amplitude of the radio signal propagating from a spacecraft to study the atmospheric and ionospheric structure of planets and satellites, planetary gravitational fields, shapes, masses, planetary rings, ephemerides of planets, solar corona, magnetic fields, cometary comae, and such aspects of the theory of general relativity as gravitational waves and gravitational redshift.

Variations in collagen fibrils network structure and surface dehydration of acid demineralized intertubular dentin: effect of dentin depth and air-exposure time.

PubMed

Fawzy, Amr S

2010-01-01

The aim was to characterize the variations in the structure and surface dehydration of acid demineralized intertubular dentin collagen network with the variations in dentin depth and time of air-exposure (3, 6, 9 and 12 min). In addition, to study the effect of these variations on the tensile bond strength (TBS) to dentin. Phosphoric acid demineralized superficial and deep dentin specimens were prepared. The structure of the dentin collagen network was characterized by AFM. The surface dehydration was characterized by probing the nano-scale adhesion force (F(ad)) between AFM tip and intertubular dentin surface as a new experimental approach. The TBS to dentin was evaluated using an alcohol-based dentin self-priming adhesive. AFM images revealed a demineralized open collagen network structure in both of superficial and deep dentin at 3 and 6 min of air-exposure. However, at 9 min, superficial dentin showed more collapsed network structure compared to deep dentin that partially preserved the open network structure. Total collapsed structure was found at 12 min for both of superficial and deep dentin. The value of the F(ad) is decreased with increasing the time of air-exposure and is increased with dentin depth at the same time of air-exposure. The TBS was higher for superficial dentin at 3 and 6 min, however, no difference was found at 9 and 12 min. The ability of the demineralized dentin collagen network to resist air-dehydration and to preserve the integrity of open network structure with the increase in air-exposure time is increased with dentin depth. Although superficial dentin achieves higher bond strength values, the difference in the bond strength is decreased by increasing the time of air-exposure. The AFM probed F(ad) showed to be sensitive approach to characterize surface dehydration, however, further researches are recommended regarding the validity of such approach.

Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists.

PubMed

Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

2013-01-01

Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior.

Deep Unsupervised Learning on a Desktop PC: A Primer for Cognitive Scientists

PubMed Central

Testolin, Alberto; Stoianov, Ivilin; De Filippo De Grazia, Michele; Zorzi, Marco

2013-01-01

Deep belief networks hold great promise for the simulation of human cognition because they show how structured and abstract representations may emerge from probabilistic unsupervised learning. These networks build a hierarchy of progressively more complex distributed representations of the sensory data by fitting a hierarchical generative model. However, learning in deep networks typically requires big datasets and it can involve millions of connection weights, which implies that simulations on standard computers are unfeasible. Developing realistic, medium-to-large-scale learning models of cognition would therefore seem to require expertise in programing parallel-computing hardware, and this might explain why the use of this promising approach is still largely confined to the machine learning community. Here we show how simulations of deep unsupervised learning can be easily performed on a desktop PC by exploiting the processors of low cost graphic cards (graphic processor units) without any specific programing effort, thanks to the use of high-level programming routines (available in MATLAB or Python). We also show that even an entry-level graphic card can outperform a small high-performance computing cluster in terms of learning time and with no loss of learning quality. We therefore conclude that graphic card implementations pave the way for a widespread use of deep learning among cognitive scientists for modeling cognition and behavior. PMID:23653617

Artificial intelligence for analyzing orthopedic trauma radiographs.

PubMed

Olczak, Jakub; Fahlberg, Niklas; Maki, Atsuto; Razavian, Ali Sharif; Jilert, Anthony; Stark, André; Sköldenberg, Olof; Gordon, Max

2017-12-01

Background and purpose - Recent advances in artificial intelligence (deep learning) have shown remarkable performance in classifying non-medical images, and the technology is believed to be the next technological revolution. So far it has never been applied in an orthopedic setting, and in this study we sought to determine the feasibility of using deep learning for skeletal radiographs. Methods - We extracted 256,000 wrist, hand, and ankle radiographs from Danderyd's Hospital and identified 4 classes: fracture, laterality, body part, and exam view. We then selected 5 openly available deep learning networks that were adapted for these images. The most accurate network was benchmarked against a gold standard for fractures. We furthermore compared the network's performance with 2 senior orthopedic surgeons who reviewed images at the same resolution as the network. Results - All networks exhibited an accuracy of at least 90% when identifying laterality, body part, and exam view. The final accuracy for fractures was estimated at 83% for the best performing network. The network performed similarly to senior orthopedic surgeons when presented with images at the same resolution as the network. The 2 reviewer Cohen's kappa under these conditions was 0.76. Interpretation - This study supports the use for orthopedic radiographs of artificial intelligence, which can perform at a human level. While current implementation lacks important features that surgeons require, e.g. risk of dislocation, classifications, measurements, and combining multiple exam views, these problems have technical solutions that are waiting to be implemented for orthopedics.

PIV-DCNN: cascaded deep convolutional neural networks for particle image velocimetry

NASA Astrophysics Data System (ADS)

Lee, Yong; Yang, Hua; Yin, Zhouping

2017-12-01

Velocity estimation (extracting the displacement vector information) from the particle image pairs is of critical importance for particle image velocimetry. This problem is mostly transformed into finding the sub-pixel peak in a correlation map. To address the original displacement extraction problem, we propose a different evaluation scheme (PIV-DCNN) with four-level regression deep convolutional neural networks. At each level, the networks are trained to predict a vector from two input image patches. The low-level network is skilled at large displacement estimation and the high- level networks are devoted to improving the accuracy. Outlier replacement and symmetric window offset operation glue the well- functioning networks in a cascaded manner. Through comparison with the standard PIV methods (one-pass cross-correlation method, three-pass window deformation), the practicability of the proposed PIV-DCNN is verified by the application to a diversity of synthetic and experimental PIV images.

Reynolds averaged turbulence modelling using deep neural networks with embedded invariance

DOE PAGES

Ling, Julia; Kurzawski, Andrew; Templeton, Jeremy

2016-10-18

There exists significant demand for improved Reynolds-averaged Navier–Stokes (RANS) turbulence models that are informed by and can represent a richer set of turbulence physics. This paper presents a method of using deep neural networks to learn a model for the Reynolds stress anisotropy tensor from high-fidelity simulation data. A novel neural network architecture is proposed which uses a multiplicative layer with an invariant tensor basis to embed Galilean invariance into the predicted anisotropy tensor. It is demonstrated that this neural network architecture provides improved prediction accuracy compared with a generic neural network architecture that does not embed this invariance property.more » Furthermore, the Reynolds stress anisotropy predictions of this invariant neural network are propagated through to the velocity field for two test cases. For both test cases, significant improvement versus baseline RANS linear eddy viscosity and nonlinear eddy viscosity models is demonstrated.« less

Reynolds averaged turbulence modelling using deep neural networks with embedded invariance

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

Ling, Julia; Kurzawski, Andrew; Templeton, Jeremy

There exists significant demand for improved Reynolds-averaged Navier–Stokes (RANS) turbulence models that are informed by and can represent a richer set of turbulence physics. This paper presents a method of using deep neural networks to learn a model for the Reynolds stress anisotropy tensor from high-fidelity simulation data. A novel neural network architecture is proposed which uses a multiplicative layer with an invariant tensor basis to embed Galilean invariance into the predicted anisotropy tensor. It is demonstrated that this neural network architecture provides improved prediction accuracy compared with a generic neural network architecture that does not embed this invariance property.more » Furthermore, the Reynolds stress anisotropy predictions of this invariant neural network are propagated through to the velocity field for two test cases. For both test cases, significant improvement versus baseline RANS linear eddy viscosity and nonlinear eddy viscosity models is demonstrated.« less

Research capacity building through North-South-South networking: towards true partnership? An exploratory study of a network for scientific support in the field of sexual and reproductive health.

PubMed

Van der Veken, K; Belaid, L; Delvaux, T; De Brouwere, V

2017-05-05

We explored the perceptions of members of the Network for Scientific Support in the field of Sexual and Reproductive Health (NetSRH) on North-South-South networking and on constraints and perspectives for South-led research. An exploratory qualitative study was conducted 18 months after the network was launched. In-depth interviews were carried out with NetSRH members (n = 15) affiliated to southern research institutions. A thematic analysis was done and N-Vivo 10 software used. A number of barriers to South-led research were identified, the most important being a lack of time, resources and research skills, and donor influence for the choice of research topics. Although the level of technical skills, such as writing proposals and scientific papers, differed among NetSRH members, all welcomed additional research capacity building. All members have deplored the lack of research management skills such as project cycle management as well as how to communicate with and get funds from donor agencies. International (local or regional) donor agencies had their own agenda with a budget already reserved for other purposes, thus priorities identified by national researchers were less taken into consideration. Systemic dependencies on external funds lead southern research partners to respond to calls for proposals mostly initiated by partners from northern institutions, leaving limited leeway for local initiatives. Southern NetSRH members perceived coaching done by the northern partners in scientific writing positively. South-South collaboration was minimal within NetSRH at this stage of the project, mainly due to time and resources constraints. NetSRH members unanimously concluded that sustainable financing of southern research centres is a necessary condition for them to initiate their own research projects. We recommend reserving funds within the international donor agencies for South-led research in order to break the vicious circle of running behind money provided by northern donors, thereby missing out on time and resources for reviewing research gaps and/or conducting needs evaluations required to initiate relevant own research.

Deepest X-Rays Ever Reveal universe Teeming With Black Holes

NASA Astrophysics Data System (ADS)

2001-03-01

For the first time, astronomers believe they have proof black holes of all sizes once ruled the universe. NASA's Chandra X-ray Observatory provided the deepest X-ray images ever recorded, and those pictures deliver a novel look at the past 12 billion years of black holes. Two independent teams of astronomers today presented images that contain the faintest X-ray sources ever detected, which include an abundance of active super massive black holes. "The Chandra data show us that giant black holes were much more active in the past than at present," said Riccardo Giacconi, of Johns Hopkins University and Associated Universities, Inc., Washington, DC. The exposure is known as "Chandra Deep Field South" since it is located in the Southern Hemisphere constellation of Fornax. "In this million-second image, we also detect relatively faint X-ray emission from galaxies, groups, and clusters of galaxies". The images, known as Chandra Deep Fields, were obtained during many long exposures over the course of more than a year. Data from the Chandra Deep Field South will be placed in a public archive for scientists beginning today. "For the first time, we are able to use X-rays to look back to a time when normal galaxies were several billion years younger," said Ann Hornschemeier, Pennsylvania State University, University Park. The group’s 500,000-second exposure included the Hubble Deep Field North, allowing scientists the opportunity to combine the power of Chandra and the Hubble Space Telescope, two of NASA's Great Observatories. The Penn State team recently acquired an additional 500,000 seconds of data, creating another one-million-second Chandra Deep Field, located in the constellation of Ursa Major. Chandra Deep Field North/Hubble Deep Field North Press Image and Caption The images are called Chandra Deep Fields because they are comparable to the famous Hubble Deep Field in being able to see further and fainter objects than any image of the universe taken at X-ray wavelengths. Both Chandra Deep Fields are comparable in observation time to the Hubble Deep Fields, but cover a much larger area of the sky. "In essence, it is like seeing galaxies similar to our own Milky Way at much earlier times in their lives," Hornschemeier added. "These data will help scientists better understand star formation and how stellar-sized black holes evolve." Combining infrared and X-ray observations, the Penn State team also found veils of dust and gas are common around young black holes. Another discovery to emerge from the Chandra Deep Field South is the detection of an extremely distant X-ray quasar, shrouded in gas and dust. "The discovery of this object, some 12 billion light years away, is key to understanding how dense clouds of gas form galaxies, with massive black holes at their centers," said Colin Norman of Johns Hopkins University. The Chandra Deep Field South results were complemented by the extensive use of deep optical observations supplied by the Very Large Telescope of the European Southern Observatory in Garching, Germany. The Penn State team obtained optical spectroscopy and imaging using the Hobby-Eberly Telescope in Ft. Davis, TX, and the Keck Observatory atop Mauna Kea, HI. Chandra's Advanced CCD Imaging Spectrometer was developed for NASA by Penn State and Massachusetts Institute of Technology under the leadership of Penn State Professor Gordon Garmire. NASA's Marshall Space Flight Center, Huntsville, AL, manages the Chandra program for the Office of Space Science, Washington, DC. TRW, Inc., Redondo Beach, California, is the prime contractor for the spacecraft. The Smithsonian's Chandra X-ray Center controls science and flight operations from Cambridge, MA. More information is available on the Internet at: http://chandra.harvard.edu AND http://chandra.nasa.gov

Automating Deep Space Network scheduling and conflict resolution

NASA Technical Reports Server (NTRS)

Johnston, Mark D.; Clement, Bradley

2005-01-01

The Deep Space Network (DSN) is a central part of NASA's infrastructure for communicating with active space missions, from earth orbit to beyond the solar system. We describe our recent work in modeling the complexities of user requirements, and then scheduling and resolving conflicts on that basis. We emphasize our innovative use of background 'intelligent' assistants' that carry out search asynchrnously while the user is focusing on various aspects of the schedule.

Publications of the Jet Propulsion Laboratory, January through December 1974. [deep space network, Apollo project, information theory, and space exploration

NASA Technical Reports Server (NTRS)

1975-01-01

Formalized technical reporting is described and indexed, which resulted from scientific and engineering work performed, or managed, by the Jet Propulsion Laboratory. The five classes of publications included are technical reports, technical memorandums, articles from the bimonthly Deep Space Network Progress Report, special publications, and articles published in the open literature. The publications are indexed by author, subject, and publication type and number.

Summary of DSN (Deep Space Network) reimbursable launch support

NASA Technical Reports Server (NTRS)

Fanelli, N. A.; Wyatt, M. E.

1988-01-01

The Deep Space Network is providing ground support to space agencies of foreign governments as well as to NASA and other agencies of the Federal government which are involved in space activities. DSN funding for support of missions other than NASA are on either a cooperative or a reimbursable basis. Cooperative funding and support are accomplished in the same manner as NASA sponsored missions. Reimbursable launch funding and support methods are described.

Two Stage Data Augmentation for Low Resourced Speech Recognition (Author’s Manuscript)

DTIC Science & Technology

2016-09-12

speech recognition, deep neural networks, data augmentation 1. Introduction When training data is limited—whether it be audio or text—the obvious...Schwartz, and S. Tsakalidis, “Enhancing low resource keyword spotting with au- tomatically retrieved web documents,” in Interspeech, 2015, pp. 839–843. [2...and F. Seide, “Feature learning in deep neural networks - a study on speech recognition tasks,” in International Conference on Learning Representations

Boosting Contextual Information for Deep Neural Network Based Voice Activity Detection

DTIC Science & Technology

2015-02-01

multi-resolution stacking (MRS), which is a stack of ensemble classifiers. Each classifier in a building block inputs the concatenation of the predictions ...a base classifier in MRS, named boosted deep neural network (bDNN). bDNN first generates multiple base predictions from different contexts of a single...frame by only one DNN and then aggregates the base predictions for a better prediction of the frame, and it is different from computationally

Building on prior knowledge without building it in.

PubMed

Hansen, Steven S; Lampinen, Andrew K; Suri, Gaurav; McClelland, James L

2017-01-01

Lake et al. propose that people rely on "start-up software," "causal models," and "intuitive theories" built using compositional representations to learn new tasks more efficiently than some deep neural network models. We highlight the many drawbacks of a commitment to compositional representations and describe our continuing effort to explore how the ability to build on prior knowledge and to learn new tasks efficiently could arise through learning in deep neural networks.

The relationship between interviewer-respondent race match and reporting of energy intake using food frequency questionnaires in the rural South United States

USDA-ARS?s Scientific Manuscript database

The purpose of the observational study was to determine whether interviewer race influences food frequency questionnaire (FFQ) reporting accuracy in a Deep South, largely African American cohort. A secondary analysis was conducted to investigate the influence of interviewer race on energy reporting ...

Deep Sequencing Reveals the Complete Genome Sequence of Sweet potato virus G from East Timor

PubMed Central

Maina, Solomon; Edwards, Owain R.; Barbetti, Martin J.; de Almeida, Luis; Ximenes, Abel

2016-01-01

We present the first complete Sweet potato virus G (SPVG) genome from sweet potato in East Timor and compare it with seven complete SPVG genomes from South Korea (three), Taiwan (two), Argentina (one), and the United States (one). It most resembles the genomes from the United States and South Korea. PMID:27609925

Alexnet Feature Extraction and Multi-Kernel Learning for Objectoriented Classification

NASA Astrophysics Data System (ADS)

Ding, L.; Li, H.; Hu, C.; Zhang, W.; Wang, S.

2018-04-01

In view of the fact that the deep convolutional neural network has stronger ability of feature learning and feature expression, an exploratory research is done on feature extraction and classification for high resolution remote sensing images. Taking the Google image with 0.3 meter spatial resolution in Ludian area of Yunnan Province as an example, the image segmentation object was taken as the basic unit, and the pre-trained AlexNet deep convolution neural network model was used for feature extraction. And the spectral features, AlexNet features and GLCM texture features are combined with multi-kernel learning and SVM classifier, finally the classification results were compared and analyzed. The results show that the deep convolution neural network can extract more accurate remote sensing image features, and significantly improve the overall accuracy of classification, and provide a reference value for earthquake disaster investigation and remote sensing disaster evaluation.

Ship detection in optical remote sensing images based on deep convolutional neural networks

NASA Astrophysics Data System (ADS)

Yao, Yuan; Jiang, Zhiguo; Zhang, Haopeng; Zhao, Danpei; Cai, Bowen

2017-10-01

Automatic ship detection in optical remote sensing images has attracted wide attention for its broad applications. Major challenges for this task include the interference of cloud, wave, wake, and the high computational expenses. We propose a fast and robust ship detection algorithm to solve these issues. The framework for ship detection is designed based on deep convolutional neural networks (CNNs), which provide the accurate locations of ship targets in an efficient way. First, the deep CNN is designed to extract features. Then, a region proposal network (RPN) is applied to discriminate ship targets and regress the detection bounding boxes, in which the anchors are designed by intrinsic shape of ship targets. Experimental results on numerous panchromatic images demonstrate that, in comparison with other state-of-the-art ship detection methods, our method is more efficient and achieves higher detection accuracy and more precise bounding boxes in different complex backgrounds.

Quicklook Constituent Abundance and Stretch Parameter Retrieval for the Juno Microwave Radiometer using Neural Networks

NASA Astrophysics Data System (ADS)

Bellotti, A.; Steffes, P. G.

2016-12-01

The Juno Microwave Radiometer (MWR) has six channels ranging from 1.36-50 cm and the ability to peer deep into the Jovian atmosphere. An Artifical Neural Network algorithm has been developed to rapidly perform inversion for the deep abundance of ammonia, the deep abundance of water vapor, and atmospheric "stretch" (a parameter that reflects the deviation from a wet adiabate in the higher atmosphere). This algorithm is "trained" by using simulated emissions at the six wavelengths computed using the Juno atmospheric microwave radiative transfer (JAMRT) model presented by Oyafuso et al. (This meeting). By exploiting the emission measurements conducted at six wavelengths and at various incident angles, the neural network can provide preliminary results to a useful precison in a computational method hundreds of times faster than conventional methods. This can quickly provide important insights into the variability and structure of the Jovian atmosphere.

Operation's Concept for Array-Based Deep Space Network

NASA Technical Reports Server (NTRS)

Bagri, Durgadas S.; Statman, Joseph I.; Gatti, Mark S.

2005-01-01

The Array-based Deep Space Network (DSNArray) will be a part of more than 10(exp 3) times increase in the downlink/telemetry capability of the Deep space Network (DSN). The key function of the DSN-Array is to provide cost-effective, robust Telemetry, Tracking and Command (TT&C) services to the space missions of NASA and its international partners. It provides an expanded approach to the use of an array-based system. Instead of using the array as an element in the existing DSN, relying to a large extent on the DSN infrastructure, we explore a broader departure from the current DSN, using fewer elements of the existing DSN, and establishing a more modern Concept of Operations. This paper gives architecture of DSN-Array and its operation's philosophy. It also describes customer's view of operations, operations management and logistics - including maintenance philosophy, anomaly analysis and reporting.

Deep Unfolding for Topic Models.

PubMed

Chien, Jen-Tzung; Lee, Chao-Hsi

2018-02-01

Deep unfolding provides an approach to integrate the probabilistic generative models and the deterministic neural networks. Such an approach is benefited by deep representation, easy interpretation, flexible learning and stochastic modeling. This study develops the unsupervised and supervised learning of deep unfolded topic models for document representation and classification. Conventionally, the unsupervised and supervised topic models are inferred via the variational inference algorithm where the model parameters are estimated by maximizing the lower bound of logarithm of marginal likelihood using input documents without and with class labels, respectively. The representation capability or classification accuracy is constrained by the variational lower bound and the tied model parameters across inference procedure. This paper aims to relax these constraints by directly maximizing the end performance criterion and continuously untying the parameters in learning process via deep unfolding inference (DUI). The inference procedure is treated as the layer-wise learning in a deep neural network. The end performance is iteratively improved by using the estimated topic parameters according to the exponentiated updates. Deep learning of topic models is therefore implemented through a back-propagation procedure. Experimental results show the merits of DUI with increasing number of layers compared with variational inference in unsupervised as well as supervised topic models.

Predicting healthcare trajectories from medical records: A deep learning approach.

PubMed

Pham, Trang; Tran, Truyen; Phung, Dinh; Venkatesh, Svetha

2017-05-01

Personalized predictive medicine necessitates the modeling of patient illness and care processes, which inherently have long-term temporal dependencies. Healthcare observations, stored in electronic medical records are episodic and irregular in time. We introduce DeepCare, an end-to-end deep dynamic neural network that reads medical records, stores previous illness history, infers current illness states and predicts future medical outcomes. At the data level, DeepCare represents care episodes as vectors and models patient health state trajectories by the memory of historical records. Built on Long Short-Term Memory (LSTM), DeepCare introduces methods to handle irregularly timed events by moderating the forgetting and consolidation of memory. DeepCare also explicitly models medical interventions that change the course of illness and shape future medical risk. Moving up to the health state level, historical and present health states are then aggregated through multiscale temporal pooling, before passing through a neural network that estimates future outcomes. We demonstrate the efficacy of DeepCare for disease progression modeling, intervention recommendation, and future risk prediction. On two important cohorts with heavy social and economic burden - diabetes and mental health - the results show improved prediction accuracy. Copyright © 2017 Elsevier Inc. All rights reserved.

DeepID-Net: Deformable Deep Convolutional Neural Networks for Object Detection.

PubMed

Ouyang, Wanli; Zeng, Xingyu; Wang, Xiaogang; Qiu, Shi; Luo, Ping; Tian, Yonglong; Li, Hongsheng; Yang, Shuo; Wang, Zhe; Li, Hongyang; Loy, Chen Change; Wang, Kun; Yan, Junjie; Tang, Xiaoou

2016-07-07

In this paper, we propose deformable deep convolutional neural networks for generic object detection. This new deep learning object detection framework has innovations in multiple aspects. In the proposed new deep architecture, a new deformation constrained pooling (def-pooling) layer models the deformation of object parts with geometric constraint and penalty. A new pre-training strategy is proposed to learn feature representations more suitable for the object detection task and with good generalization capability. By changing the net structures, training strategies, adding and removing some key components in the detection pipeline, a set of models with large diversity are obtained, which significantly improves the effectiveness of model averaging. The proposed approach improves the mean averaged precision obtained by RCNN [16], which was the state-of-the-art, from 31% to 50.3% on the ILSVRC2014 detection test set. It also outperforms the winner of ILSVRC2014, GoogLeNet, by 6.1%. Detailed component-wise analysis is also provided through extensive experimental evaluation, which provides a global view for people to understand the deep learning object detection pipeline.

Self-Paced Prioritized Curriculum Learning With Coverage Penalty in Deep Reinforcement Learning.

PubMed

Ren, Zhipeng; Dong, Daoyi; Li, Huaxiong; Chen, Chunlin; Zhipeng Ren; Daoyi Dong; Huaxiong Li; Chunlin Chen; Dong, Daoyi; Li, Huaxiong; Chen, Chunlin; Ren, Zhipeng

2018-06-01

In this paper, a new training paradigm is proposed for deep reinforcement learning using self-paced prioritized curriculum learning with coverage penalty. The proposed deep curriculum reinforcement learning (DCRL) takes the most advantage of experience replay by adaptively selecting appropriate transitions from replay memory based on the complexity of each transition. The criteria of complexity in DCRL consist of self-paced priority as well as coverage penalty. The self-paced priority reflects the relationship between the temporal-difference error and the difficulty of the current curriculum for sample efficiency. The coverage penalty is taken into account for sample diversity. With comparison to deep Q network (DQN) and prioritized experience replay (PER) methods, the DCRL algorithm is evaluated on Atari 2600 games, and the experimental results show that DCRL outperforms DQN and PER on most of these games. More results further show that the proposed curriculum training paradigm of DCRL is also applicable and effective for other memory-based deep reinforcement learning approaches, such as double DQN and dueling network. All the experimental results demonstrate that DCRL can achieve improved training efficiency and robustness for deep reinforcement learning.

75 FR 63786 - Fisheries of the Caribbean, Gulf of Mexico, and South Atlantic; Reef Fish Fishery of the Gulf of...

Federal Register 2010, 2011, 2012, 2013, 2014

2010-10-18

... reduce the commercial quota for gag and, thus, the combined commercial quota for shallow-water grouper... IFQ account holder's deep-water grouper (DWG) allocation has been landed and sold, or transferred, or... percent of their gross revenue in 2008 and 2009 respectively. Revenue from deep-water grouper (DWG...

Douglas-fir survival and growth in response to spring planting date and depth

Treesearch

R. O. Strothmann

1971-01-01

Douglas-fir seedlings were planted by four methods in late February and late March on hot, south-facing slopes in northwestern California. Besides standard planting, the techniques used included deep planting at two different depths, and shading the lower stem. The differences in survival after 3 growing seasons were not statistically significant, but deep planting had...

76 FR 78879 - Fisheries of the Caribbean, Gulf of Mexico, and South Atlantic; Snapper-Grouper Fishery Off the...

Federal Register 2010, 2011, 2012, 2013, 2014

2011-12-20

... implemented, this rule would remove the harvest and possession prohibition of six deep-water snapper-grouper... intent of this rule is to reduce the socio-economic impacts to fishermen harvesting deep-water snapper... obtained from the Southeast Regional Office Web site at http://sero.nmfs.noaa.gov . FOR FURTHER INFORMATION...

The past, present and future distribution of a deep-sea shrimp in the Southern Ocean

PubMed Central

Costello, Mark J.

2016-01-01

Shrimps have a widespread distribution across the shelf, slope and seamount regions of the Southern Ocean. Studies of Antarctic organisms have shown that individual species and higher taxa display different degrees of sensitivity and adaptability in response to environmental change. We use species distribution models to predict changes in the geographic range of the deep-sea Antarctic shrimp Nematocarcinus lanceopes under changing climatic conditions from the Last Glacial Maximum to the present and to the year 2100. The present distribution range indicates a pole-ward shift of the shrimp population since the last glaciation. This occurred by colonization of slopes from nearby refugia located around the northern part of Scotia Arc, southern tip of South America, South Georgia, Bouvet Island, southern tip of the Campbell plateau and Kerguelen plateau. By 2100, the shrimp are likely to expand their distribution in east Antarctica but have a continued pole-ward contraction in west Antarctica. The range extension and contraction process followed by the deep-sea shrimp provide a geographic context of how other deep-sea Antarctic species may have survived during the last glaciation and may endure with projected changing climatic conditions in the future. PMID:26925334