Search NASA⌕ Search

SEARCH · Search NASA

Results for “machine learning efficiency”

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.

At least 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Synthesis challenges, thermodynamic stability, and growth kinetics of La–Si–P ternary compounds

Although many new compounds have been recently predicted with the help of machine learning, the successful experimental synthesis of these compounds remains challenging. Computational insights about the thermodynamic stability and phase formation kinetics among the ground state and competing metastable phases are highly desirable to rationalize and attempt to overcome synthesis challenges experimentally. In this work, we explore synthetic challenges within ternary La–Si–P compounds through feedback between experimental and computational studies. We discuss the experimental challenges in forming three computationally predicted ternary phases (La 2 SiP, La 5 SiP 3 , and La 2 SiP 3 ). To understand the synthetic challenges, we performed molecular dynamics (MD) simulations using an accurate and efficient artificial neural network machine learning (ANN-ML) interatomic potential. We study the phase stability and formation kinetics of these ternary phases in relation to the reported and synthesized La 2 SiP 4 phase. While the growth of the La 2 SiP 4 phase can be reproduced by our MD simulation, our results indicate that the rapid formation of a Si-substituted LaP crystalline phase is a major barrier to the synthesis of the predicted La 2 SiP, La 5 SiP 3 , and La 2 SiP 3 ternary compounds, agreeing well with experimental observations. Our simulations also suggest that there is a narrow temperature window in which the La 2 SiP 3 phase can be grown from the solid–liquid interface.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Efficient sampling of constrained high-dimensional theoretical spaces with machine learning

Abstract Models of physics beyond the Standard Model often contain a large number of parameters. These form a high-dimensional space that is computationally intractable to fully explore. Experimental results project onto a subspace of parameters that are consistent with those observations, but mapping these constraints to the underlying parameters is also typically intractable. Instead, physicists often resort to scanning small subsets of the full parameter space and testing for experimental consistency. We propose an alternative approach that uses generative models to significantly improve the computational efficiency of sampling high-dimensional parameter spaces. To demonstrate this, we sample the constrained and phenomenological Minimal Supersymmetric Standard Models subject to the requirement that the sampled points are consistent with the measured Higgs boson mass. Our method achieves orders of magnitude improvements in sampling efficiency compared to a brute force search.

Hollingsworth, Jacob↗
DOE OSTI2025

Accurate and efficient predictions of keyhole dynamics in laser materials processing using machine learning-aided simulations

The keyhole phenomenon has been widely observed in laser materials processing, including laser welding, remelting, cladding, drilling, and additive manufacturing. Keyhole-induced defects, primarily pores, dramatically affect the performance of final products, impeding the broad use of these laser-based technologies. The formation of these pores is typically associated with the dynamic behavior of the keyhole. So far, the accurate characterization and prediction of keyhole features, particularly keyhole depth, as a function of time, has been a challenging task. In situ characterization of keyhole dynamic behavior using the synchrotron X-ray technique is informative but complicated and expensive. Current simulations are generally hindered by their poor accuracy and generalization abilities in predicting keyhole depths due to the lack of accurate laser absorptance data. In this study, we develop a machine learning-aided simulation method that accurately predicts keyhole dynamics, especially in keyhole depth fluctuations, over a wide range of processing parameters. In two case studies involving titanium and aluminum alloys, we achieve keyhole depth prediction with a mean absolute percentage error of 10 %, surpassing those simulated using the ray-tracing method with an error margin of 30 %, while also reducing computational time. This exceptional fidelity and efficiency empower our model to serve as a cost-effective alternative to synchrotron experiments. Our machine learning-aided simulation method is affordable and readily deployable for a large variety of materials, opening new doors to eliminate or reduce defects for a wide range of laser materials processing techniques.

Computational fluid dynamics↗
DOE OSTI2026

Measuring the Energy Consumption and Efficiency of Deep Neural Networks: An Empirical Analysis and Design Recommendations

Addressing the "Red-AI" trend of rising energy consumption by large-scale neural networks, this study investigates the measured energy consumption of training various fully connected neural network architectures. We introduce the BUTTER-E dataset, an augmentation to the BUTTER Empirical Deep Learning dataset, containing energy consumption and performance data from 41,129 individual experimental runs spanning 30,582 distinct configurations: 13 datasets, 20 sizes (trainable parameters), 8 "shapes", and 14 depths on both CPUs and GPUs using node-level watt-meters. This dataset reveals the complex relationship between dataset size, network structure, and energy use. Our analysis uncovers a surprising, hardware-mediated non-linear relationship between energy efficiency and network design, challenging the assumption that reducing the number of parameters or FLOPs is the best way to achieve greater energy efficiency. We propose a straightforward and effective energy model that accounts for network size, computing, and memory hierarchy. Highlighting the need for cache-considerate algorithm development, we suggest a codesign approach to energy efficient network, algorithm, and hardware design. This work contributes to the fields of sustainable computing and Green AI, offering practical guidance for creating more energy-efficient neural networks and promoting sustainable AI.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Machine Learning for Anomaly Detection in Neural Network Security and SRF Cavities

This dissertation explores the development and deployment of machine learning approaches to address critical challenges in anomaly detection across two distinct domains: neural network security in federated learning settings and cavity behavior analysis in particle accelerator operations at Jefferson Lab in Newport News, Virginia. Anomaly detection identifies deviations from expected patterns, safeguarding systems in cybersecurity, industry, and research against malicious activities and failures. This dissertation demonstrates how our machine learning approaches enhance detection accuracy and efficiency in both neural network security and industrial applications. First, we investigate vulnerabilities in deep neural networks deployed in federated learning. Although federated learning preserves user privacy by training models locally, it remains vulnerable to backdoor attacks, in which malicious participants embed hidden triggers that induce targeted misbehavior. We propose a self-supervised contrastive learning framework to detect and mitigate such backdoor attacks. In our experiments, this method achieves higher detection accuracy and lower false positive rates than existing defenses, while operating without access to local model updates or original training data and thus preserving the privacy guarantees of the federated setting. Second, we address the operational reliability of superconducting radio-frequency (SRF) cavities at the Continuous Electron Beam Accelerator Facility (CEBAF). Our research leverages an unsupervised learning approach, combined with Principal Component Analysis (PCA) and k-means clustering, to identify anomalous behaviors in SRF cavities. Our method detects subtle anomalous behavior by analyzing SRF signal data. This knowledge allows for the early detection and resolution of potential faults, significantly improving the efficiency and reliability of operations. Third, we extend these insights to time-series anomaly detection more broadly. We design a contrastive-learning based model tailored to increasingly dynamic environments and academic research. This model improves detection accuracy in settings that require real-time monitoring and predictive maintenance. Our research underscores the broader applicability and impact of advanced machine learning techniques in anomaly detection. By extracting meaningful patterns from complex data, machine learning can significantly enhance security in distributed neural networks and improve the efficiency of particle accelerator operations. This dissertation serves as a stepping stone for future investigations into the vast possibilities of anomaly detection, inspiring further exploration and development of machine learning techniques in this field.

Ferguson, Hal [Old Dominion University]↗
DOE OSTI2024

New Directions for Thermoelectrics: A Roadmap from High‐Throughput Materials Discovery to Advanced Device Manufacturing

Thermoelectric materials, which can convert waste heat into electricity or act as solid‐state Peltier coolers, are emerging as key technologies to address global energy shortages and environmental sustainability. However, discovering materials with high thermoelectric conversion efficiency is a complex and slow process. The emerging field of high‐throughput material discovery demonstrates its potential to accelerate the development of new thermoelectric materials combining high efficiency and low cost. The synergistic integration of high‐throughput material processing and characterization techniques with machine learning algorithms can form an efficient closed‐loop process to generate and analyze broad datasets to discover new thermoelectric materials with unprecedented performances. Meanwhile, the recent development of advanced manufacturing methods provides exciting opportunities to realize scalable, low‐cost, and energy‐efficient fabrication of thermoelectric devices. This review provides an overview of recent advances in discovering thermoelectric materials using high‐throughput methods, including processing, characterization, and screening. Advanced manufacturing methods of thermoelectric devices are also introduced to realize the broad impacts of thermoelectric materials in power generation and solid‐state cooling. In the end, this article also discusses the future research prospects and directions.

Song, Kaidong↗
DOE OSTI2025

Dynamical Sketching for Enhanced Communication Efficiency in Federated Learning

Federated learning (FL) has revolutionized distributed machine learning by enabling collaborative model training without sharing local data. However, communication efficiency and privacy guarantees remain significant challenges. This paper introduces a dynamic sketching mechanism in FL, optimizing the trade-off between communication efficiency and model accuracy. By dynamically selecting the sketch matrix size, our approach adapts to the evolving characteristics of the data and the model, ensuring optimal performance across diverse scenarios. We leverage Bayesian optimization to systematically tune the sketch parameters, achieving an effective balance between resource efficiency and model performance. Experimental results on the MNIST dataset using a convolutional neural network (CNN) architecture validate the proposed method's efficiency and scalability. Our dynamic sketching approach significantly outperforms fixed-size sketching techniques, achieving higher compression ratios (up to 62x) and providing better privacy guarantees while maintaining high model accuracy. These findings highlight the robustness and versatility of our approach and make it a valuable solution for privacy-preserving, communication-efficient federated learning.

Afrose, Sharmin [ORNL]↗
DOE OSTI2024

Efficient graph representation framework for chemical molecule similarity tasks

Graph data has emerged in numerous scientific domains and machine learning techniques have been widely used for analysis and learning of diverse data for prediction and decision. Machine learning techniques can readily address complex problems by leveraging their structural information. But graphs cannot be directly used for existing machine learning algorithms unless encoded as vectors. The problem of efficient representation of graphs is a substantial challenge in graph machine learning. In this paper, we propose a novel two-stage framework for the representation of chemical molecule graphs based on the strengths of Graph Isomorphism Networks (GINs) and Siamese autoencoders. In the first stage, the GIN model is constructed and trained using the structural information of chemical molecule graphs. Node attributes, edge attributes, and edge indices are used as input data, while graph attributes are used as labels. The GIN model effectively captures the structural characteristics of graphs and can accurately predict graph attributes, i.e., molecular properties. It also generates Graph Embeddings, represented as vectors that encode the structural information of graphs. In the second stage, Graph Embedding vectors are further optimized for downstream similarity tasks while preserving the graph structural information. The Siamese autoencoder is constructed and trained, which reduces the dimensionality of the Graph Embedding vectors, while maximizing the preservation of structural information in the original high-dimensional vectors. The resulting low-dimensional Graph Embeddings can be effectively utilized for tasks such as approximate nearest neighbor search. The experimental results demonstrate the effectiveness of our proposed framework in accurately predicting graph similarity.

Ma, Jiaji↗
DOE OSTI2020

Machine-Learning-Driven, Site-Specific Weather Forecasting for Grid-Interactive Efficient Buildings: Preprint

Emerging grid-interactive efficient buildings (GEBs) have great potential to provide much-needed demand flexibility to electric grids while fulfilling their own control targets by co-optimizing smart appliances, solar photovoltaics, electric vehicles, and energy storage at buildings. To enable the optimal operation of GEBs, site-specific weather information—such as temperature, solar irradiance, relative humidity, and wind speed—is crucial; however, this information is generally unavailable or expensive to obtain. This paper develops advanced machine learning methods to provide precise weather forecasts for individual building sites using readily available weather station data. Support vector regression and artificial neural networks have been employed to learn the spatiotemporal correlations between the weather conditions at nearby weather stations and the individual building site. The proposed site-specific weather forecasting methods have been validated using 1-year actual weather measurement data collected in the Denver metro area. Results show that the developed machine-learning-driven methods can accurately forecast the temperature at the target building site 1 hour ahead with mean absolute error less than 0.72°C and a 48% improvement over the persistence method. Site-specific weather forecasts will improve the understanding of the microclimate effect and its impact on building energy consumption. This information will drive efficiency upgrades and adjustments of building control strategies to improve energy savings and increase flexibility in building loads.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2023

Spatiotemporal Super-Resolution with Generative Machine Learning for Creating Renewable Energy Resource Data Under Climate Change Scenarios

As we plan for a future with higher penetrations of renewables and increasing electrification, it becomes more important to understand how the electricity grid will operate under a variety of weather events. We must also consider that the weather our future grid will experience will be different and possibly more extreme than the historical weather that we have extensive data for. We can use data from global climate models (GCMs) to help understand how our climate may change over the next several decades, but there is often a significant gap between the low-resolution GCM data and the high-resolution weather data required to study power systems under specific weather events. Therefore, our objective in this work is to develop tools that can bridge this gap by using low-resolution GCM data to create realistic high-resolution weather datasets that can be used to study renewable energy generation and electricity demand. To accomplish this objective, we have developed a set of generative machine learning models that can rapidly downscale GCM daily average output data at an approximate grid resolution of 100km to hourly data at an approximate 4 km grid resolution. The models can be used to create high resolution data from nearly any GCM included in the Coupled Model Intercomparison Project (CMIP) Phase 5 or 6. Our methods include all datasets regularly used to study the integration of wind and solar power plants as well as changes in electricity demand due to heating and cooling loads. These models and datasets enable power systems modelers to study climate change-influenced weather events and their impact on the grid. We have downscaled and validated wind, solar, temperature, and humidity data with very promising results. The generative machine learning methods are computationally efficient and produce data that has similar statistical characteristics to current state-of-the-art historical datasets. We have trained initial generative models and produced an initial dataset collectively referred to as Sup3rCC: Super-Resolved Renewable Energy Resource Data with Climate Change Impacts. The data covers a (mostly) historical period from 2015-2025 and a future period from 2050-2059. We have also taken hypothetical high-electrification load data and scaled the heating and cooling loads with respect to the 2050-2059 high-resolution Sup3rCC meteorology. The results show how future levels of renewable energy generation and electrified load may be impacted by climate change, setting the stage for capacity expansion models to consider a dynamic climate through model years.

climate change↗
DOE OSTI2024

A Hybrid Anomaly Detection Approach for Obfuscated Malware

With the rapid evolution of malicious software, cyber threats have become increasingly sophisticated, employing advanced obfuscation techniques to evade traditional detection methods. This study presents a hybrid anomaly detection approach applied to obfuscated malware. Even though there is a large body of research in this field, existing malware detection techniques have some drawbacks, such as requiring large amounts of data, trustworthiness (imprecise results) of algorithms, and advanced obfuscation. To overcome these challenges, there is a need to employ solid and efficient techniques for malware detection. This paper proposes a hybrid approach, combining an autoencoder with traditional machine-learning methods to create an efficient malware detection framework. We used the malware memory dataset (MalMemAnalysis-2022) to evaluate this framework. The results indicate that our proposed approach can detect obfuscated malware when a deep autoencoder used for feature learning is combined with logistic regression, and it is extremely fast with an Accuracy, Detection Rate (DR), Matthew Correlation Coefficient(MCC), and Statistical Parity Difference

malware detection, Hybrid Anomly Detection, Obfusc↗
DOE OSTI2021

Improve Learning from Crowds via Generative Augmentation

Crowdsourcing provides an efficient label collection schema for supervised machine learning. However, to control annotation cost, each instance in the crowdsourced data is typically annotated by a small number of annotators. This creates a sparsity issue and limits the quality of machine learning models trained on such data. In this paper, we study how to handle sparsity in crowdsourced data using data augmentation. Specifically, we propose to directly learn a classifier by augmenting the raw sparse annotations. We implement two principles of high-quality augmentation using Generative Adversarial Networks: 1) the generated annotations should follow the distribution of authentic ones, which is measured by a discriminator; 2) the generated annotations should have high mutual information with the ground-truth labels, which is measured by an auxiliary network. Extensive experiments and comparisons against an array of state-of-the-art learning from crowds methods on three real-world datasets proved the effectiveness of our data augmentation framework. It shows the potential of our algorithm for low-budget crowdsourcing in general.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2022

Efficient hybrid explicit-implicit learning for multiscale problems

Splitting method is a powerful method to handle application problems by splitting physics, scales, domain, and so on. Many splitting algorithms have been designed for efficient temporal discretization. Here, in this paper, our goal is to use temporal splitting concepts in designing machine learning algorithms and, at the same time, help splitting algorithms by incorporating data and speeding them up. We propose a machine learning assisted splitting scheme which improves the efficiency of the scheme meanwhile preserves the accuracy. We consider a recently introduced multiscale splitting algorithms, where the multiscale problem is solved on a coarse grid. To approximate the dynamics, only a few degrees of freedom are solved implicitly, while others explicitly. This splitting concept allows identifying degrees of freedom that need implicit treatment. In this paper, we use this splitting concept in machine learning and propose several strategies. First, the implicit part of the solution can be learned as it is more difficult to solve, while the explicit part can be computed. This provides a speed-up and data incorporation for splitting approaches. Secondly, one can design a hybrid neural network architecture because handling explicit parts requires much fewer communications among neurons and can be done efficiently. Thirdly, one can solve the coarse grid component via PDEs or other approximation methods and construct simpler neural networks for the explicit part of the solutions. We discuss these options and implement one of them by interpreting it as a machine translation task. This interpretation of the splitting scheme successfully enables us using the Transformer since it can perform model reduction for multiple time series and learn the connection between them. We also find that the splitting scheme is a great platform to predict the coarse solution with insufficient information of the target model: the target problem is partially given and we need to solve it through a known problem which approximates the target. Our machine learning model can incorporate and encode the given information from two different problems and then solve the target problems. We conduct four numerical examples and the results show that our method is stable and accurate.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Deep learning techniques to disentangle water use efficiency, climate change, and carbon sequestration across ecosystem scales

Focal Areas: We plan to use machine learning (ML) to disentangle the impact of climate change on plant water use efficiency (WUE) across the leaf level, structured canopy, and community ecosystem scales. Due to the complex behavior of modeling plant WUE’s connection to carbon storage (and hidden co-varying interactions) we suggest that physics-based deep learning must be utilized. The second topic we propose is, based on a better understanding of WUE response to climate change, there is a need to predict how WUE will limit or enhance the recycling of water from transpiration by vegetation back to the land surface worldwide, and thus impact water storage capacity (i.e., natural and reservoirs), using machine learning optimization techniques such as genetic algorithms.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Machine learning approach for vibronically renormalized electronic band structures

Here, we present a machine learning (ML) method for efficient computation of vibrational thermal expectation values of physical properties from first principles. Our approach is based on the nonperturbative frozen phonon formulation in which stochastic Monte Carlo algorithm is employed to sample configurations of nuclei in a supercell at finite temperatures based on a first-principles phonon model. A deep-learning neural network is trained to accurately predict physical properties associated with sampled phonon configurations, thus bypassing the time-consuming ab initio calculations. To incorporate the point-group symmetry of the electronic system into the ML model, group-theoretical methods are used to develop a symmetry-invariant descriptor for phonon configurations in the supercell. We apply our ML approach to compute the temperature dependent electronic energy gap of silicon based on density functional theory (DFT). We show that, with less than a hundred DFT calculations for training the neural network model, an order of magnitude larger number of sampling can be achieved for the computation of the vibrational thermal expectation values. Our work highlights the promising potential of ML techniques for finite temperature first-principles electronic structure methods.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

A machine learning based approach to online electron reconstruction at CLAS12

Online reconstruction is key for monitoring purposes and real time analysis in High Energy and Nuclear Physics experiments. A necessary component of reconstruction algorithms is particle identification that combines information left by a particle passing through several detector components to identify the particle’s type. Of particular interest to electro-production Nuclear Physics experiments such as CLAS12 is electron identification which is used to trigger data recording. A machine learning approach was developed for CLAS12 to reconstruct and identify electrons by combining raw signals at the data acquisition level from several detector components. Here, this approach achieves an electron identification purity above 75% whilst retaining an efficiency close to 100%. The machine learning tools are capable of running at high rates exceeding the data acquisition rates and will allow electron reconstruction in real-time. This work enhances online analyses and monitoring and can contribute to improved triggering at CLAS12. This machine learning driven approach will also be crucial for experiments aiming to transition to streaming readout operations where online reconstruction will be a key component of the data taking paradigm.

Artificial intelligence↗
DOE OSTI2025

FIRM image analysis: A machine learning workflow for quantifying extracellular matrix components from electron microscopy images

The extracellular matrix (ECM) is a complex network of biomolecules that plays an integral role in the structure, processes, and signaling mechanisms of cells and tissues. Identifying and quantifying changes in these matrix components provides insight into the mechanisms behind specific tissue remodeling processes; however, quantifying these changes is challenging due to difficult imaging conditions, complexity of the ECM, and the subtlety of these changes. Current imaging techniques allow us to visualize these critical remodeling events and developments in image analysis have employed a combination of analysis software and machine learning techniques to improve the efficiency and accuracy with which features are measured. Although image analysis has seen much improvement in recent years, there has been no technique developed to address ambiguity in feature edges in electron microscopy images. Presented here is a new machine learning-based workflow for the analysis of microscopy images named FIRM (Feature Identification from Raw Microscopy) that uses a random forest classifier to identify ECM features of interest and generate binary segmentation masks for quantification with ImageJ-FIJI. FIRM performed with an F1 score of 0.794 and greater than 80% accuracy for number and size of features detected. FIRM had similar deviation from the ground truth in the number of identified fibrils, fibril size, and size distributions when compared to human analyses. The results suggest that FIRM performs as well as manual analysis and requires a fraction of the time. This analysis technique is more efficient, eliminates user bias, and can be easily optimized to identify a variety of features, making it useful for any discipline requiring image analysis.

Science & Technology - Other Topics↗
DOE OSTI2024

Integrating Machine Learning Potential and X-ray Absorption Spectroscopy for Predicting the Chemical Speciation of Disordered Carbon Nitrides

Precise determination of atomic structural information in functional materials holds transformative potential and broad implications for emerging technologies. Spectroscopic techniques, such as X-ray absorption near-edge structure (XANES), have been widely used for material characterization; however, extracting chemical information from experimental probes remains a significant challenge, particularly for disordered materials. We present an integrated approach that combines atomic simulations, data-driven techniques, and experimental measurements to investigate chemical speciation of amorphous carbon nitride systems as a case study. Here, we discuss the development of machine learning potentials that can efficiently explore the vast configuration space of amorphous carbon nitrides. By employing statistical methods, this structural database enables the elucidation of the most representative local structures and how they evolve with chemical compositions and density. Density functional theory simulations are used to establish a correlation between the local structure and spectroscopic signatures, which then serve as the basis for interpreting and extracting chemical content from experimental data. Although our framework is specifically demonstrated for XANES and carbon nitrides, the approach described herein is readily adaptable as applied to other experimental characterization probes and materials classes.

36 MATERIALS SCIENCE↗