Learning constitutive relations from indirect observations using deep neural networks
Not provided.
Computer Science↗
SEARCH · Search NASA
Results for “neural network 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.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Quadrature Based Neural Network Learning of Stochastic Hamiltonian Systems
Hamiltonian Neural Networks (HNNs) provide structure-preserving learning of Hamiltonian systems. In this paper, we extend HNNs to structure-preserving inversion of stochastic Hamiltonian systems (SHSs) from observational data. We propose the quadrature-based models according to the integral form of the SHSs’ solutions, where we denoise the loss-by-moment calculations of the solutions. The integral pattern of the models transforms the source of the essential learning error from the discrepancy between the modified Hamiltonian and the true Hamiltonian in the classical HNN models into that between the integrals and their quadrature approximations. This transforms the challenging task of deriving the relation between the modified and the true Hamiltonians from the (stochastic) Hamilton–Jacobi PDEs, into the one that only requires invoking results from the numerical quadrature theory. Meanwhile, denoising via moments calculations gives a simpler data fitting method than, e.g., via probability density fitting, which may imply better generalization ability in certain circumstances. Numerical experiments validate the proposed learning strategy on several concrete Hamiltonian systems. The experimental results show that both the learned Hamiltonian function and the predicted solution of our quadrature-based model are more accurate than that of the corrected symplectic HNN method on a harmonic oscillator, and the three-point Gaussian quadrature-based model produces higher accuracy in long-time prediction than the Kramers–Moyal method and the numerics-informed likelihood method on the stochastic Kubo oscillator as well as other two stochastic systems with non-polynomial Hamiltonian functions. Moreover, the Hamiltonian learning error εH arising from the Gaussian quadrature-based model is lower than that from Simpson’s quadrature-based model. These demonstrate the superiority of our approach in learning accuracy and long-time prediction ability compared to certain existing methods and exhibit its potential to improve learning accuracy via applying precise quadrature formulae.
Mathematics↗
ACES-GNN: can graph neural network learn to explain activity cliffs?
Graph Neural Networks (GNNs) have revolutionized molecular property prediction by leveraging graph-based representations, yet their opaque decision-making processes hinder broader adoption in drug discovery. This study introduces the Activity-Cliff-Explanation-Supervised GNN (ACES-GNN) framework, designed to simultaneously improve predictive accuracy and interpretability by integrating explanation supervision for activity cliffs (ACs) into GNN training. ACs, defined by structurally similar molecules with significant potency differences, pose challenges for traditional models due to their reliance on shared structural features. By aligning model attributions with chemist-friendly interpretations, the ACES-GNN framework bridges the gap between prediction and explanation. Validated across 30 pharmacological targets, ACES-GNN consistently enhances both predictive accuracy and attribution quality for ACs compared to unsupervised GNNs. Our results demonstrate a positive correlation between improved predictions and accurate explanations, offering a robust and adaptable framework to better understand and interpret ACs. This work underscores the potential of explanation-guided learning to advance interpretable artificial intelligence in molecular modeling and drug discovery.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
Do graph neural networks learn traditional jet substructure?
At the CERN LHC, the task of jet tagging, whose goal is to infer the origin of a jet given a set of final-state particles, is dominated by machine learning methods. Graph neural networks have been used to address this task by treating jets as point clouds with underlying, learnable, edge connections between the particles inside. We explore the decision-making process for one such state-of-the-art network, ParticleNet, by looking for relevant edge connections identified using the layerwise-relevance propagation technique. As the model is trained, we observe changes in the distribution of relevant edges connecting different intermediate clusters of particles, known as subjets. The resulting distribution of subjet connections is different for signal jets originating from top quarks, whose subjets typically correspond to its three decay products, and background jets originating from lighter quarks and gluons. This behavior indicates that the model is using traditional jet substructure observables, such as the number of prongs -- energetic particle clusters -- within a jet, when identifying jets.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
Learning generative neural networks with physics knowledge
Deep generative neural networks have enabled modeling complex distributions, but incorporating physics knowledge into the neural networks is still challenging and is at the core of current physics-based machine learning research. To this end, we propose a physics generative neural network (PhysGNN), a new class of generative neural networks for learning unknown distributions in a physical system described by partial differential equations (PDE). PhysGNN couples PDE systems with generative neural networks. It is a fully differentiable model that allows back-propagation of gradients through both numerical PDE solvers and generative neural networks, and is trained by minimizing the discrete Wasserstein distance between generated and observed probability distributions of the PDE outputs using the stochastic gradient descent method. Moreover, PhysGNN does not require adversarial training like standard generative neural networks, which offers better stability than adversarial training. We show that PhysGNN can learn complex distributions in stochastic inverse problems, where conventional methods such as maximum likelihood estimation and momentum matching methods may be inapplicable when little knowledge is known about the form of unknown distributions or the physical model is too complex. Furthermore, our method allows physics-based generative neural network training for learning complex distributions in the context of differential equations.
97 MATHEMATICS AND COMPUTING↗
Comparative Studies of the Structural and Transport Properties of Molten Salt FLiNaK Using the Machine-Learned Neural Network and Reparametrized Classical Forcefields
Despite surging interest in molten salt reactors and thermal storage systems, knowledge of the physicochemical properties of molten salts are still inadequate due to demanding experiments that require high temperature, impurity control, and corrosion mitigation. Therefore, the ability to predict these properties for molten salts from first-principles computations is urgently needed. Herein, we developed and compared a machine-learned neural network force field (NNFF) and a reparametrized rigid ion model (RIM) for a prototypical molten salt LiF–NaF–KF (FLiNaK). We found that NNFF was able to reproduce both the structural and transport properties of the molten salt with first-principles accuracy and classical-MD computational efficiency. Furthermore, the correlation between the local atomic structures and the dynamics was identified by comparing with RIMs, suggesting the significance of polarization of anions implicitly embedded in the NNFF. Furthermore, this work demonstrated a computational framework that can facilitate the screening of molten salts with different chemical compositions, impurities, and additives, and at different thermodynamic conditions suitable for the next-generation nuclear reactors and thermal energy storage facilities.
36 MATERIALS SCIENCE↗
Development of interatomic potential for Al–Tb alloys using a deep neural network learning method
An interatomic potential for the Al–Tb alloy around the composition of Al 90 Tb 10 is developed using the deep neural network (DNN) learning method. The atomic configurations and the corresponding total potential energies and forces on each atom obtained from ab initio molecular dynamics (AIMD) simulations are collected to train a DNN model to construct the interatomic potential for the Al–Tb alloy. Here we show that the obtained DNN model can well reproduce the energies and forces calculated by AIMD simulations. Molecular dynamics (MD) simulations using the DNN interatomic potential also accurately describe the structural properties of the Al 90 Tb 10 liquid, such as partial pair correlation functions (PPCFs) and bond angle distributions, in comparison with the results from AIMD simulations. Furthermore, the developed DNN interatomic potential predicts the formation energies of the crystalline phases of the Al–Tb system with an accuracy comparable to ab initio calculations. The structure factors of the Al 90 Tb 10 metallic liquid and glass obtained by MD simulations using the developed DNN interatomic potential are also in good agreement with the experimental X-ray diffraction data. The development of short-range order (SRO) in the Al 90 Tb 10 liquid and the undercooled liquid is also analyzed and three dominant SROs, i.e., Al-centered distorted icosahedron (DISICO) and Tb-centered ‘3661’ and ‘15551’ clusters, respectively, are identified.
36 MATERIALS SCIENCE↗
Deep neural network learning of complex binary sorption equilibria from molecular simulation data
We employed deep neural networks (NNs) as an efficient and intelligent surrogate of molecular simulations for complex sorption equilibria using probabilistic modeling.
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
A Transductive Graph Neural Network learning for Grid Resilience Analysis
Power grids are critical infrastructures that require robust resilience analysis to ensure reliable and uninterrupted electricity supply. Traditional simulation-based methods for grid resilience analysis suffer from computational complexity and limited ability to capture the full spectrum of potential disruptions. This paper presents a novel approach to enhance grid resilience by leveraging transductive graph neural network (GNN) learning to identify critical nodes and links. By leveraging the graph structure and system features, GNNs effectively learn resilience metrics and accurately identify critical nodes based on actual grid operational behavior. The efficacy of the proposed approach is demonstrated through case studies on node criticality scoring and critical node/line identification in cascading outage scenarios. The results highlight the advantages of learning-based methods over traditional simulation-based approaches and their potential to revolutionize grid resilience analysis. The contributions of this paper include a graph-based scalable approach for fast cascading analysis, an inductive formulation for training GNN models, and a transfer learning-based approach to scale the model to largescale power systems.
grid resilience, graph neural networks, transducti↗
Neural-network learning of SPOD latent dynamics
Here, we aim to reconstruct the latent space dynamics of high dimensional, quasi-stationary systems using model order reduction via the spectral proper orthogonal decomposition (SPOD). The proposed method is based on three fundamental steps: in the first, once that the mean flow field has been subtracted from the realizations (also referred to as snapshots), we compress the data from a high-dimensional representation to a lower dimensional one by constructing the SPOD latent space; in the second, we build the time-dependent coefficients by projecting the snapshots containing the fluctuations onto the SPOD basis and we learn their evolution in time with the aid of recurrent neural networks; in the third, we reconstruct the high-dimensional data from the learnt lower -dimensional representation. The proposed method is demonstrated on two different test cases, namely, a compressible jet flow, and a geophysical problem known as the Madden-Julian Oscillation. An extensive comparison between SPOD and the equivalent POD-based counterpart is provided and differences between the two approaches are highlighted. The numerical results suggest that the proposed model is able to provide low rank predictions of complex statistically stationary data and to provide insights into the evolution of phenomena characterized by specific range of frequencies. The comparison between POD and SPOD surrogate strategies highlights the need for further work on the characterization of the interplay of error between data reduction techniques and neural network forecasts.
97 MATHEMATICS AND COMPUTING↗
Monitoring of Liquid Metal Reactor Heater Zones with Recurrent Neural Network Learning of Temperature Time Series
Advanced high-temperature fluid reactors (ARs), such as sodium fast reactors (SFRs) and molten salt cooled reactors (MSCRs) utilize high-temperature fluids at ambient pressure. To melt the fluid during reactor startup and prevent fluid freezing during cooldown, the thermal–hydraulic systems of such ARs include heater zones consisting of specific heaters with controllers, temperature sensors, and thermal insulation. The failure of heater zones due to insulation material degradation or improper installation, resulting in parasitic heat losses, can lead to fluid freezing. The detection of faults using a heat-transfer model is difficult because of a lack of knowledge of the experimental details. Data-driven machine learning of heater zone temperature time series offers a viable alternative. In this study, we benchmarked the performance of recurrent neural networks (RNNs) in an analysis of heat-up transient temperature time series of heater zones installed on a liquid sodium vessel. The RNN models include long short-term memory (LSTM) and gated recurrent unit (GRU) networks, as well as their bi-directional variants, BiLSTM and BiGRU. Anomalous temperature points were designated using a percentile-based threshold applied to residual fluctuations in the detrended temperature time series. Additionally, the impact of the exponentially weighted moving average (EWMA) method on detection accuracy was examined. The RNN models’ performance was assessed using precision, recall, and F 1 score metrics. Results demonstrated that RNN models effectively detect anomalies in temperature time series with the best models for each heater zone achieving F 1 scores of over 93%. To explain the variations in RNN model performance across different heater zones, we used Kullback–Leibler (KL) divergence to quantify the relative entropy between training and testing data, and the Detrended Fluctuation Analysis (DFA) to assess long-range temporal correlations. For datasets with strong long-range correlations and minimal relative entropy between training and testing data, GRU is the best-performing model. When the data exhibits weaker long-term correlations and a significant relative entropy between training and testing distributions, BiGRU shows the best performance. For the data sets with intermediate values of both KL divergence and DFA, the best performance is obtained with LSTM and BiLSTM, respectively.
gated recurrent unit↗
Electron energy loss spectroscopy database synthesis and automation of core-loss edge recognition by deep-learning neural networks
Abstract The ionization edges encoded in the electron energy loss spectroscopy (EELS) spectra enable advanced material analysis including composition analyses and elemental quantifications. The development of the parallel EELS instrument and fast, sensitive detectors have greatly improved the acquisition speed of EELS spectra. However, the traditional way of core-loss edge recognition is experience based and human labor dependent, which limits the processing speed. So far, the low signal–noise ratio and the low jump ratio of the core-loss edges on the raw EELS spectra have been challenging for the automation of edge recognition. In this work, a convolutional-bidirectional long short-term memory neural network (CNN-BiLSTM) is proposed to automate the detection and elemental identification of core-loss edges from raw spectra. An EELS spectral database is synthesized by using our forward model to assist in the training and validation of the neural network. To make the synthesized spectra resemble the real spectra, we collected a large library of experimentally acquired EELS core edges. In synthesize the training library, the edges are modeled by fitting the multi-Gaussian model to the real edges from experiments, and the noise and instrumental imperfectness are simulated and added. The well-trained CNN-BiLSTM network is tested against both the simulated spectra and real spectra collected from experiments. The high accuracy of the network, 94.9%, proves that, without complicated preprocessing of the raw spectra, the proposed CNN-BiLSTM network achieves the automation of core-loss edge recognition for EELS spectra with high accuracy.
36 MATERIALS SCIENCE↗
A Self-learning Neural Network Approach for Radio Frequency Interference Detection and Removal in Radio Astronomy
In this work, we present a novel neural network (NN) method for the detection and removal of radio frequency interference (RFI) from the raw digitized signal in the signal processing chain of a typical radio astronomy experiment. The main advantage of our method is that it does not require a training set. Instead, our method relies on the fact that the true signal of interest coming from astronomical sources is thermal and therefore described as a Gaussian random process, which cannot be compressed. We employ a variational encoder/decoder network to find the compressible information in the data stream that can explain the most variance with the fewest degrees of freedom. We demonstrate it on a set of toy problems and stored ring buffers from the Baryon Mapping eXperiment prototype. We find that the RFI subtraction is effective at cleaning simulated time streams: While we find that the power spectra of the RFI-cleaned time streams output by the NN suffer from extra signal consistent with additive noise, we find that it is generally around percent level across the band and sub 10% in contaminated spectral channels even when RFI power is an order of magnitude larger than the signal. We discuss the advantages and limitations of this method and possible implementation in the front end of future radio experiments.
79 ASTRONOMY AND ASTROPHYSICS↗
Practical database enrichment strategies with iterative learning: Neural Network Potentials for Phase Change Memory.
Abstract not provided.
Adams, David↗
Evaluation of Time-step Frequency on Prediction Accuracy Applied to Deep Learning Neural Network Surrogate Models for CO2 Storage
This PowerPoint presentation was presented virtually at the American Geophysical Union (AGU) Fall Meeting 2020.
Vikara, Derek↗
Efficient transfer learning for neural network language models.
Abstract not provided.
Skryzalin, Jacek↗
Scalable Causal Graph Learning through a Deep Neural Network
Learning the causal graph in a complex system is crucial for knowledge discovery and decision making, yet it remains a challenging problem because of the unknown nonlinear interaction among system components. Most of the existing methods either rely on predefined kernel or data distribution, or they focus simply on the causality between a single target and the remaining system. This work presents a deep neural network for scalable causal graph learning (SCGL) through low-rank approximation. The SCGL model can explore nonlinearity on both temporal and intervariable relationships without any predefined kernel or distribution assumptions. Through low-rank approximation, the noise influence is reduced, and better accuracy and high scalability are achieved. Experiments using synthetic and real-world datasets show that our SCGL algorithm outperforms existing state-of-the-art methods for causal graph learning.
97 MATHEMATICS AND COMPUTING↗