SEARCH · Search NASA

Results for “limits of machine learning”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 883 records · Page 49

All sources NASA NTRS DOE OSTI

DOE OSTI2024

Entropy removal of medical diagnostics

Shannon entropy is a core concept in machine learning and information theory, particularly in decision tree modeling. To date, no studies have extensively and quantitatively applied Shannon entropy in a systematic way to quantify the entropy of clinical situations using diagnostic variables (true and false positives and negatives, respectively). Decision tree representations of medical decision-making tools can be generated using diagnostic variables found in literature and entropy removal can be calculated for these tools. This concept of clinical entropy removal has significant potential for further use to bring forth healthcare innovation, such as quantifying the impact of clinical guidelines and value of care and applications to Emergency Medicine scenarios where diagnostic accuracy in a limited time window is paramount. This analysis was done for 623 diagnostic tools and provided unique insights into their utility. For studies that provided detailed data on medical decision-making algorithms, bootstrapped datasets were generated from source data to perform comprehensive machine learning analysis on these algorithms and their constituent steps, which revealed a novel and thorough evaluation of medical diagnostic algorithms.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Improving the reliability of machine learned potentials for modeling inhomogeneous liquids

The atomic-scale response of inhomogeneous fluids at interfaces and surrounding solute particles plays a critical role in governing chemical, electrochemical, and biological processes. Classical molecular dynamics simulations have been applied extensively to simulate the response of fluids to inhomogeneities directly, but are limited by the accuracy of the underlying interatomic potentials. Here, we use neural network potentials (NNPs) trained to ab initio simulations to accurately predict the inhomogeneous responses of two distinct fluids: liquid water and molten NaCl. Although NNPs can be readily trained to model complex bulk systems across a range of state points, we show that to appropriately model a fluid's response at an interface, relevant inhomogeneous configurations must be included in the training data. In order to sufficiently sample appropriate configurations of such inhomogeneous fluids, we develop protocols based on molecular dynamics simulations in the presence of external potentials. We demonstrate that NNPs trained on inhomogeneous fluid configurations can more accurately predict several key properties of fluids—including the density response, surface tension and size-dependent cavitation free energies—for liquid water and molten NaCl, compared to both empirical interatomic potentials and NNPs that are not trained on such inhomogeneous configurations. This work therefore provides a first demonstration and framework to extract the response of inhomogeneous fluids from first principles for classical density-functional treatment of fluids free from empirical potentials.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

Weighted active space protocol for multireference machine-learned potentials

Multireference methods such as multiconfiguration pair-density functional theory accurately capture electronic correlation in systems with strong multiconfigurational character, but their cost precludes direct use in molecular dynamics. Combining these methods with machine-learned interatomic potentials (MLPs) can extend their reach. However, the sensitivity of multireference calculations to the choice of the active space complicates the consistent evaluation of energies and gradients across structurally diverse nuclear configurations. To overcome this limitation, we introduce the weighted active space protocol (WASP), a systematic approach to assign a consistent active space for a given system across uncorrelated configurations. By integrating WASP with MLPs and enhanced sampling techniques, we propose a data-efficient active learning cycle that enables the training of an MLP on multireference data. We demonstrated the approach on the TiC + -catalyzed C–H activation of methane, a reaction that poses challenges for Kohn–Sham density functional theory due to its significant multireference character. This framework enables accurate and efficient modeling of catalytic dynamics, establishing a paradigm for simulating complex reactive processes beyond the limits of conventional electronic-structure methods.

enhanced sampling↗

DOE OSTI2022

A machine learning approach to determine the elastic properties of printed fiber-reinforced polymers

This work focuses on the simultaneous determination of the elastic constants and the fiber orientation state for a short fiber-reinforced polymer composite by performing a minimum of experimental tests. Here we introduce a methodology that enables the inverse determination of fiber orientation state and the in-situ polymer properties by performing tensile tests at the composite coupon level. We demonstrate the approach for the extrusion deposition additive manufacturing (EDAM) process to illustrate one application of the methodology, but the development is such that it can be applied to short fiber-reinforced polymer (SFRP) systems processed via other methods. Currently, developing composites additive manufacturing digital twins require extensive material characterization. In particular, the mechanical characterization of the orthotropic elastic properties of a composite involves extensive sample preparation and testing, therefore the elasticity tensor is generally populated using a micromechanics model. This, however, requires measuring the fiber orientation state in addition to knowing the constituent material properties. Experimentally measuring the fiber orientation state can be tedious and time consuming. Further, optical methods are limited to resolving the orientation of cylindrical fibers or cluster of non-cylindrical fibers, and computed tomography (CT) methods scan regions of volume that are much smaller than a full printed bead. Therefore, we propose a methodology, accelerated by machine learning, to identify the anisotropic mechanical properties and fiber orientation state at the same time. Early results show that inference of the fiber orientation and composite properties is possible with as few as three tensile tests. Our results show that a combination of the choice of the micromechanics model and reliable set of experiments can yield the nine elastic constants, as well as, the fiber orientation state.

36 MATERIALS SCIENCE↗

NASA NTRSDate not supplied

Hybrid Model Based Approaches for Systems Health Management and Prognostics

To facilitate and solve the prediction problem, awareness of the current health state of the system is key, since it is necessary to perform condition-based predictions. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Latest achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision making. In principle, data driven approaches can replace expensive experimental test-setups as well as reduce the number of simulations needed to explore, e.g., the parametric space of a multi-parameter model. Nonetheless, the limitations of pure data-driven methods came to light rather quickly, at least for some industries. In many industrial applications, data acquisition is costly, and the volume of data that can be collected does not satisfy the requirements for an effective model training and cross-validation. Therefore, some recent works in the area of machine learning is focusing on blending physics with data-driven algorithms, thus mitigating the drawbacks of the two approaches and emphasizing respective advantages. Partial physical knowledge of the problem can aid the learning process by “guiding” the algorithm towards efficient solutions that satisfy the physics driving the system behavior. The result is a hybrid modeling approach combining physical knowledge as well data driven methods to develop a unified hybrid approach. A hybrid framework for fusing information from physics-based performance models along with deep learning algorithms for prognostics of complex safety critical systems is presented. In this framework, physics-based performance models infer unobservable model parameters related to the system's components health solving a calibration problem in the deep learning approach.

Hybrid Modeling↗

NASA NTRSDate not supplied

Hybrid Approaches to Systems Health Management and Prognostics

Systems Health Management↗

DOE OSTI2014

Finding Waldo: Learning about Users from their Interactions

Visual analytics is inherently a collaboration between human and computer. However, in current visual analytics systems, the computer has limited means of knowing about its users and their analysis processes. While existing research has shown that a user’s interactions with a system reflect a large amount of the user’s reasoning process, there has been limited advancement in developing automated, real-time techniques that mine interactions to learn about the user. In this paper, we demonstrate that we can accurately predict a user’s task performance and infer some user personality traits by using machine learning techniques to analyze interaction data. Specifically, we conduct an experiment in which participants perform a visual search task and we apply well-known machine learning algorithms to three encodings of the users interaction data. We achieve, depending on algorithm and encoding, between 62% and 96% accuracy at predicting whether each user will be fast or slow at completing the task. Beyond predicting performance, we demonstrate that using the same techniques, we can infer aspects of the user’s personality factors, including locus of control, extraversion, and neuroticism. Further analyses show that strong results can be attained with limited observation time, in some cases, 82% of the final accuracy is gained after a quarter of the average task completion time. Overall, our findings show that interactions can provide information to the computer about its human collaborator, and establish a foundation for realizing mixed- initiative visual analytics systems.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2022

Hybrid Model Based Approaches for Systems Health Management and Prognostics

This is a previously approved and published presentation. To accurately predict the future state of any system, it is required to possess knowledge of its current health state and future operational conditions. Present achievements of data-driven algorithms in regression of complex nonlinear functions and classification tasks have generated a growing interest in artificial intelligence for industrial applications. Complex multi-physics models as well as digital twins, once purely built on physics and corresponding simplified lumped parameter iterations, can now benefit from machine learning algorithms to mitigate the lack of understanding of some complex behavior. Given models of the current and future system behavior, a general approach of model-based prognostics can solve the prediction problem and further decision-making. In principle, data-driven approaches can replace expensive experimental test-setups as well as reduce the number of simulations needed to explore, e.g., the parametric space of a multi-parameter model. Nonetheless, the limitations of pure data-driven methods came to light rather quickly, at least for some industries. In many industrial applications, data acquisition is costly, and the volume of data that can be collected does not satisfy the requirements for effective model training and cross-validation. Therefore, some recent works in the area of machine learning is focusing on blending physics with data-driven algorithms, thus mitigating the drawbacks of the two approaches and emphasizing respective advantages. Partial physical knowledge of the problem can aid the learning process by “guiding” the algorithm towards efficient solutions that satisfy the physics driving the system behavior. The result is a hybrid modeling approach combining physical knowledge as well data-driven methods to develop a unified hybrid approach. A hybrid framework for fusing information from physics-based performance models along with deep learning algorithms for prognostics of complex safety-critical systems is presented. In this framework, physics-based performance models infer unobservable model parameters related to the system's components health solving a calibration problem in the deep learning approach.

Prognostics↗

DOE OSTI2018

DeepHyper: Asynchronous Hyperparameter Search for Deep Neural Networks

Hyperparameters employed by deep learning (DL) methods play a substantial role in the performance and reliability of these methods in practice. Unfortunately, finding performance-optimizing hyperparameter settings is a notoriously difficult task. Hyperparameter search methods typically have limited production-strength implementations or do not target scalability within a highly parallel machine, portability across different machines, experimental comparison between different methods, and tighter integration with workflow systems. In this paper, we present DeepHyper, a Python package that provides a common interface for the implementation and study of scalable hyperparameter search methods. It adopts the Balsam workflow system to hide the complexities of running large numbers of hyperparameter configurations in parallel on high-performance computing (HPC) systems. We implement and study asynchronous model-based search methods that consist of sampling a small number of input hyperparameter configurations and progressively fitting surrogate models over the input-output space until exhausting a user-defined budget of evaluations. We evaluate the efficacy of these methods relative to approaches such as random search, genetic algorithms, Bayesian optimization, and hyperband on DL benchmarks on CPU- and GPU-based HPC systems.

Balaprakash, Prasanna↗

DOE OSTI2021

TQEA: Temporal Quantum Error Analysis

For the last few years quantum computers have captured researchers’ interest in solving complicated problems in finance, chemistry, machine learning, and many other NP problems. Quantum computing benefits from parallel computation, entanglement, and calculation speedup compared to classical computing. Even though the operations and number of qubits in quantum computers are limited, previous research showed that using quantum computers can be beneficial. These benefits gained researchers’ interest to work on the improvement of quantum computers and overcome the limitations by building more advanced quantum computers with fault tolerance capability, more physical qubits, and better operational gates. In practice, Quantum applications would be translated into Quantum Circuits and each circuit can contain one or many entanglements (s), single qubits operations, and readouts. The importance of running Quantum applications with controlled errors or without errors is critical to the desired result from Quantum readout. In this paper, we study the characteristics of superconducting quantum machines along with the previously proposed methods of Quantum error detection/correction and in order to build a better error model we propose a temporal method for analyzing the the behavior of qubits errors (T1 and T2), frequency of operation, entanglement errors, and readout errors. In order to detect and predict the behavior of each machine, we use simple exponential smoothing analysis on heuristic calibration data. Finally, based on observation, we show that based on frequency, qubits, readout and single operation errors we can build a temporal based model to study and predict the outcome of runs on each Quantum machine.

Baheri, Betis↗

DOE OSTI2024

Complexity-calibrated benchmarks for machine learning reveal when prediction algorithms succeed and mislead

Abstract Recurrent neural networks are used to forecast time series in finance, climate, language, and from many other domains. Reservoir computers are a particularly easily trainable form of recurrent neural network. Recently, a “next-generation” reservoir computer was introduced in which the memory trace involves only a finite number of previous symbols. We explore the inherent limitations of finite-past memory traces in this intriguing proposal. A lower bound from Fano’s inequality shows that, on highly non-Markovian processes generated by large probabilistic state machines, next-generation reservoir computers with reasonably long memory traces have an error probability that is at least $$\sim 60\%$$ ∼ 60 % higher than the minimal attainable error probability in predicting the next observation. More generally, it appears that popular recurrent neural networks fall far short of optimally predicting such complex processes. These results highlight the need for a new generation of optimized recurrent neural network architectures. Alongside this finding, we present concentration-of-measure results for randomly-generated but complex processes. One conclusion is that large probabilistic state machines—specifically, large $$\epsilon$$ ϵ -machines—are key to generating challenging and structurally-unbiased stimuli for ground-truthing recurrent neural network architectures.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2020

Correlation-based network analysis combined with machine learning techniques highlight the role of the GABA shunt in Brachypodium sylvaticum freezing tolerance

Perennial grasses will account for approximately 16 billion gallons of renewable fuels by the year 2022, contributing significantly to carbon and nitrogen sequestration. However, perennial grasses productivity can be limited by severe freezing conditions in some geographical areas, although these risks could decrease with the advance of climate warming, the possibility of unpredictable early cold events cannot be discarded. We conducted a study on the model perennial grass Brachypodium sylvaticum to investigate the molecular mechanisms that contribute to cold and freezing adaption. The study was performed on two different B. sylvaticum accessions, Ain1 and Osl1, typical to warm and cold climates, respectively. Both accessions were grown under controlled conditions with subsequent cold acclimation followed by freezing stress. For each treatment a set of morphological parameters, transcription, metabolite, and lipid profiles were measured. State-of-the-art algorithms were employed to analyze cross-component relationships. Phenotypic analysis revealed higher adaption of Osl1 to freezing stress. Our analysis highlighted the differential regulation of the TCA cycle and the GABA shunt between Ain1 and Osl1. Osl1 adapted to freezing stress by repressing the GABA shunt activity, avoiding the detrimental reduction in fatty acid biosynthesis and the concomitant detrimental effects on membrane integrity.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2018

SuperNeurons: Dynamic GPU Memory Management for Training Deep Neural Networks

Going deeper and wider in neural architectures improves their accuracy, while the limited GPU DRAM places an undesired restriction on the network design domain. Deep Learning (DL) practitioners either need to change to less de- sired network architectures, or nontrivially dissect a network across multiGPUs. These distract DL practitioners from concentrating on their original machine learning tasks. We present SuperNeurons: a dynamic GPU memory scheduling runtime to enable the network training far beyond the GPU DRAM capacity. SuperNeurons features 3 memory optimizations, Liveness Analysis, Unified Tensor Pool, and Cost-Aware Recomputation; together they effectively reduce the network-wide peak memory usage down to the maximal memory usage among layers. We also address the performance issues in these memory-saving techniques. Given the limited GPU DRAM, SuperNeurons not only provisions the necessary memory for the training, but also dynamically allocates the memory for convolution workspaces to achieve the high performance. Evaluations against Caffe, Torch, MXNet and TensorFlow have demonstrated that SuperNeurons trains at least 3.2432 deeper network than current ones with the leading performance. Particularly, SuperNeurons can train ResNet2500 that has 104 basic network layers on a 12GB K40c.

Wang, Linnan↗

DOE OSTI2024

Offshore application of landslide susceptibility mapping using gradient-boosted decision trees: a Gulf of Mexico case study

Abstract Among natural hazards occurring offshore, submarine landslides pose a significant risk to offshore infrastructure installations attached to the seafloor. With the offshore being important for current and future energy production, there is a need to anticipate where future landslide events are likely to occur to support planning and development projects. Using the northern Gulf of Mexico (GoM) as a case study, this paper performs Landslide Susceptibility Mapping (LSM) using a gradient-boosted decision tree (GBDT) model to characterize the spatial patterns of submarine landslide probability over the United States Exclusive Economic Zone (EEZ) where water depths are greater than 120 m. With known spatial extents of historic submarine landslides and a Geographic Information System (GIS) database of known topographical, geomorphological, geological, and geochemical factors, the resulting model was capable of accurately forecasting potential locations of sediment instability. Results of a permutation modelling approach indicated that LSM accuracy is sensitive to the number of unique training locations with model accuracy becoming more stable as the number of training regions was increased. The influence that each input feature had on predicting landslide susceptibility was evaluated using the SHapely Additive exPlanations (SHAP) feature attribution method. Areas of high and very high susceptibility were associated with steep terrain including salt basins and escarpments. This case study serves as an initial assessment of the machine learning (ML) capabilities for producing accurate submarine landslide susceptibility maps given the current state of available natural hazard-related datasets and conveys both successes and limitations.

Dyer, Alec S. (ORCID:0000000219813904)↗

DOE OSTI2025

Harnessing distributed GPU computing for generalizable graph convolutional networks in power grid reliability assessments

Although machine learning (ML) has emerged as a powerful tool for rapidly assessing grid contingencies, prior studies have largely considered a static grid topology in their analyses. This limits their application, since they need to be re-trained for every new topology. Here, this paper explores the development of generalizable graph convolutional network (GCN) models by pre-training them across a range of grid topologies and contingency types. We found that a GCN model with auto-regressive moving average (ARMA) layers with a line graph representation of the grid offered the best predictive performance in predicting voltage magnitudes (VM) and voltage angles (VA). We introduced the concept of phantom nodes to consider disparate grid topologies with a varying number of nodes and lines. For pre-training the GCN ARMA model across a variety of topologies, distributed graphics processing unit (GPU) computing afforded us significant training scalability. The predictive performance of this model on grid topologies that were part of the training data is substantially better than the direct current (DC) approximation. Although direct application of the pre-trained model to topologies that are not part of the grid is not particularly satisfactory, fine-tuning with small amounts of data from a specific topology of interest significantly improves predictive performance. In general, this paper highlights the feasibility of training large-scale GNN models to assess the reliability of power grids by considering a wide variety of grid topologies and contingency types. With the advent of foundational models in ML and the exponential increase in GPU computing clusters, generalizable ML models will significantly enhance how utilities manage power systems and make decisions in real-time or near-real-time.

24 - POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2023

Search for nonresonant Higgs boson pair production in final state with two bottom quarks and two tau leptons in proton-proton collisions at $\sqrt{s}$ = 13 TeV

A search for the nonresonant production of Higgs boson pairs (HH ) via gluon-gluon and vector boson fusion processes in final states with two bottom quarks and two tau leptons is presented. The search uses data from proton-proton collisions at a center-of-mass energy of $\sqrt{s}$ = 13 TeV recorded with the CMS detector at the LHC, corresponding to an integrated luminosity of 138 fb -1 . Events in which at least one tau lepton decays hadronically are considered and multiple machine learning techniques are used to identify and extract the signal. The data are found to be consistent, within uncertainties, with the standard model (SM) predictions. Upper limits on the HH production cross section are set to constrain the parameter space for anomalous Higgs boson couplings. The observed (expected) upper limit at 95% confidence level corresponds to 3.3 (5.2) times the SM prediction for the inclusive HH cross section and to 124 (154) times the SM prediction for the vector boson fusion HH cross section. At 95% confidence level, the Higgs field self-coupling is constrained to be within -1.7 and 8.7 times the SM expectation, and the coupling of two Higgs bosons to two vector bosons is constrained to be within -0.4 and 2.6 times the SM expectation.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2019

BraggNet: integrating Bragg peaks using neural networks

Neutron crystallography offers enormous potential to complement structures from X-ray crystallography by clarifying the positions of low-Z elements, namely hydrogen. Macromolecular neutron crystallography, however, remains limited, in part owing to the challenge of integrating peak shapes from pulsed-source experiments. To advance existing software, this article demonstrates the use of machine learning to refine peak locations, predict peak shapes and yield more accurate integrated intensities when applied to whole data sets from a protein crystal. The artificial neural network, based on the U-Net architecture commonly used for image segmentation, is trained using about 100 000 simulated training peaks derived from strong peaks. After 100 training epochs (a round of training over the whole data set broken into smaller batches), training converges and achieves a Dice coefficient of around 65%, in contrast to just 15% for negative control data sets. Integrating whole peak sets using the neural network yields improved intensity statistics compared with other integration methods, including k-nearest neighbours. These results demonstrate, for the first time, that neural networks can learn peak shapes and be used to integrate Bragg peaks. In conclusion, it is expected that integration using neural networks can be further developed to increase the quality of neutron, electron and X-ray crystallography data.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

Anomaly detection for MPC forecast in Fleet of Water Heaters

Among residential devices, water heaters consume 20% of home energy use in the United States. Water heaters possess the capability to store energy within their reservoirs, enabling the ability to decouple energy use from hot water use. This capability can be used to reduce energy usage and costs while also supporting grid services. This requires accurate forecasting of the parameters of the water heater such as upper and lower temperatures. In this study, we analyzed the performance and behavior of a water heater model used in the real-world to predict a control mechanism that is implemented in a smart residential neighborhood. The model forecasts are accurate in most cases but not all. In such scenarios, error correction of the model is necessary to further improve model predictive control accuracy. Anomaly detection is the first step of error correction. This study complements existing research by grouping time series data into two clusters one with anomalies and another without anomalies. To achieve this task, we explored and compared multiple unsupervised machine learning algorithms to perform clustering. Among these algorithms, Ward clustering has the lowest running time and identified the highest number of anomalies for the upper temperature limit. The proposed approach is tested based on the data collected in a neighborhood with 46 townhomes located in Atlanta, GA.

Lebakula, Viswadeep↗