SEARCH · Search NASA

Results for “interpretability”

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 397 records · Page 22

All sources NASA NTRS DOE OSTI

DOE OSTI2025

Advanced Photovoltaic Module Characterization: Using Image Transformers for Current–Voltage Curve Prediction From Electroluminescence Images

Individual photovoltaic (PV) module health monitoring can be a daunting task for operation and maintenance of solar farms. Modules can be inspected through luminescence, thermal imaging, and current–voltage (I–V) curve analyzes for identification of damage and power loss. I–V curves provide easily interpretable data to determine module health as they directly provide electrical performance metrics. However, in order to obtain these curves, modules must be disconnected from the array and either removed to a solar simulator or characterized in situ with corrections for module temperature, the incident solar spectrum, and intensity. Luminescence or thermal images of a module are relatively easy to acquire in situ. Electroluminescence (EL) images highlight physical defects in the modules but do not provide easily interpretable features to correlate with electrical performance. This work presents a SWin transformer network to predict I–V curves for PV modules from their corresponding EL images. The predicted I–V curves allow the accurate prediction of the maximum power point (MPP), short-circuit current I sc , and open-circuit voltage V oc with a mean error less of than 1%. Comparing single diode model (SDM) parameters extracted from the predicted curves to those extracted from the true curves, the series resistance R s demonstrates a mean error of 5.19%, and the photocurrent I a mean error of 0.197%. The shunt resistance R sh and dark current Io parameters are predicted with larger errors because of their sensitivity to small changes in the I–V curve.

Byford, Brandon K. [New Mexico State Univ., Las Cr↗

DOE OSTI2026

Practical Three-Dimensional Curvilinear Aperture Monopulse Synthetic Aperture Radar

When collected over a curved path, Synthetic Aperture Radar (SAR) contains information about the three-dimensional content of a scene that can aid interpretation of this difficult imaging modality. In a point scatterer simulation with amplitudes and heights derived from a real SAR image pair, we supplement confocal processing with a monopulse-based algorithm to compute heights of certain bright points, which we find automatically. In SAR, bright points often reflect man-made infrastructure central to scene interpretation.

Story, Mark [ORNL] (ORCID:0000000289148911)↗

DOE OSTI2024

Deep Koopman Neural Network for Analyzing High-Energy-Density Simulations of Electrical Wire Explosions

Megaampere-scale electrical wire experiments (EWEs) provide a platform for studying magnetohydrodynamic (MHD) instability growth in magneto-inertial fusion (MIF) devices. Even when nonlinear simulations of these experiments can digitally reproduce much of the experimentally observed instability growth, interpreting the results and understanding mode growth and evolution can be non-trivial. As a first step toward providing better interpretation of these simulation features, this work investigates the use of a deep neural network that uses Koopman operator theory to analyze the dynamics of pulsed-power-driven explosions of EWEs. This deep neural network is trained on 1-D resistive MHD simulations of EWEs. This neural network learns to transform the nonlinear data into a lower-dimensional representation where the time dynamics are linear. Layers of this neural network are shown to learn features of the simulations, including the locations of shock waves and different physical regimes of the simulation. Using the learned features, the network can compress a time state of the simulation consisting of 5120 data point into a 36-parameter lower-dimensional latent space embedding. Furthermore, these embeddings are shown to be clustered in the latent space by initial radius and time state.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DOE OSTI2024

Visual Analytics of Multivariate Networks With Representation Learning and Composite Variable Construction

Multivariate networks are commonly found in real-world data-driven applications. Uncovering and understanding the relations of interest in multivariate networks is not a trivial task. This article presents a visual analytics workflow for studying multivariate networks to extract associations between different structural and semantic characteristics of the networks (e.g., what are the combinations of attributes largely relating to the density of a social network?). The workflow consists of a neural-network-based learning phase to classify the data based on the chosen input and output attributes, a dimensionality reduction and optimization phase to produce a simplified set of results for examination, and finally an interpreting phase conducted by the user through an interactive visualization interface. A key part of our design is a composite variable construction step that remodels nonlinear features obtained by neural networks into linear features that are intuitive to interpret. We demonstrate the capabilities of this workflow with multiple case studies on networks derived from social media usage and also evaluate the workflow with qualitative feedback from experts.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

SigTime: Learning and Visually Explaining Time Series Signatures

Understanding and distinguishing temporal patterns in time series data is essential for scientific discovery and decision-making. For example, in biomedical research, uncovering meaningful patterns in physiological signals can improve diagnosis, risk assessment, and patient outcomes. However, existing methods for time series pattern discovery face major challenges, including high computational complexity, limited interpretability, and difficulty in capturing meaningful temporal structures. Here, to address these gaps, we introduce a novel learning framework that jointly trains two Transformer models using complementary time series representations: shapelet-based representations to capture localized temporal structures and traditional feature engineering to encode statistical properties. The learned shapelets serve as interpretable signatures that differentiate time series across classification labels. Additionally, we develop a visual analytics system—SigTime—with coordinated views to facilitate exploration of time series signatures from multiple perspectives, aiding in useful insights generation. We quantitatively evaluate our learning framework on eight publicly available datasets and one proprietary clinical dataset. Additionally, we demonstrate the effectiveness of our system through two usage scenarios along with the domain experts: one involving public ECG data and the other focused on preterm labor analysis.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2025

On the bulk compaction of brittle granular materials, part I: SeS analysis of axial compression to 4000 MPa

The bulk compaction of granular materials has been studied for decades to interpret and manage responses for soils and powder-based component fabrication, and geophysical, celestial, and ballistic impact. Their bulk or macroscopic compaction response is limited by what occurs at the granular or microstructural scale. Motivation existed to more closely examine that association specific to granular brittle materials (e.g., ceramics and glasses). That examination is offered in a series of three companion papers where Part I describes a new supplemental analysis adopted to bulk compaction response involving relatively high compaction stresses (4000 MPa). Bulk compactions of vitreous silicates and crystalline quartzes were interpreted in three ways, including that of a new analysis that considers the product of void ratio (e) and stress (S) as a function of S, hereafter referred to as “SeS analysis”. In conclusion, the SeS analysis was found to be an informative supplement to conventional bulk compaction analyses because it provides more consistent higher sensitivity for the identification of bulk density rate increase with increasing compaction (softening); a rate increase that arises from the cumulative effect of the onsets and progression of compaction-induced yielding, fracture or comminution, densification, phase change, or combinations thereof occurring at the granular or microstructural scale.

compaction↗

DOE OSTI2024

The Driving Force Controls Charge Separation Distance in Solid-State Organic Donor/Acceptors

In this work, charge pairs of varying separation distance are interpreted from photoinduced absorption detected magnetic resonance (PADMR) spectroscopy of organic, small molecule dilute donor/acceptor thin films. We report that a donor/acceptor film that generates few free charges at room temperature yet has a relatively large, calculated driving force for electron transfer generates a large concentration of tightly bound CT states when measured with PADMR. These states are markedly absent in films with smaller driving forces yet higher free charge yields which instead only show charge-separated state signals with weaker spin coupling. We interpret this result to be in support of a hypothesis where a larger reorganization energy associated with charge transfer to tightly bound CT states means that they are primarily generated in systems far from the Marcus optimum for free charge yield. And the highest free charge yielding systems instead predominantly undergo long-range charge separation into the acceptor host.

charge separation↗

DOE OSTI2024

Insights into the thermo-hydraulic properties of compacted MX80 bentonite during hydration under elevated temperature

In high-level radioactive waste geological repositories, compacted bentonite undergoes coupled thermo-hydraulic processes due to heat released from a central waste canister and groundwater imbibition from the surrounding host rock. An understanding of these processes is essential for long-term simulations of radionuclide migration and canister corrosion, which requires an understanding of temperature effects on the coupled thermo-hydraulic properties governing these processes. In this study, a tank-scale radial infiltration test was used to investigate water imbibition processes in compacted bentonite under a central heater temperature of 200 °C that simulates high thermal gradients in a repository. Interpretation of this test focuses on evaluation of the liquid water wetting front during hydration and the interpretation of the temperature-dependent transient soil water retention curve (SWRC), thermal conductivity function (TCF), and hydraulic conductivity function (HCF). The SWRC during imbibition follows a temperature-dependent wetting path. Temperature effects on the hydraulic conductivity of bentonite in saturated conditions had the greatest effect on the shape of the HCF, with minimal temperature effects at higher suctions. The transient thermal conductivity data matched well with a new TCF linked with the SWRC shape. Although the bentonite layer was restrained, local deformations during hydration may have affected the shapes of the TCF and HCF.

Engineering↗

DOE OSTI2024

Large $$N_c$$ QCD phase diagram at $$\mu _B=0$$

Abstract Lattice studies suggest that at zero baryon chemical potential and increasing temperature there are three characteristic regimes in QCD that are connected by smooth analytical crossovers: a hadron gas regime at$$T < T_{ch}\sim 155$$ T < T ch ∼ 155 MeV, an intermediate regime, called stringy fluid, at$$T_{ch}< T < \sim 3 T_{ch}$$ T ch < T < ∼ 3 T ch , and a quark-gluon plasma regime at higher temperatures. These regimes have been interpreted to reflect different approximate symmetries and effective degrees of freedom. In the hadron gas the effective degrees of freedom are hadrons and the approximate chiral symmetry of QCD is spontaneously broken. The intermediate regime has been interpreted as lacking spontaneous chiral symmetry breaking along with the emergence of new approximate symmetry, chiral spin symmetry, that is not a symmetry of the Dirac Lagrangian, but is a symmetry of the confining part of the QCD Lagrangian. While the high temperature regime is the usual quark-gluon plasma which is often considered to reflect “deconfinement” in some way. This paper explores the behavior of these regimes of QCD as the number of colors in the theory,$$N_c$$ N c , gets large. In the large$$N_c$$ N c limit the theory is center-symmetric and notions of confinement and deconfinement are unambiguous. The energy density is$$\mathcal{O}(N_c^0)$$ O ( N c 0 ) in the meson gas,$${{\mathcal {O}}}(N_c^1)$$ O ( N c 1 ) in the intermediate regime and$${{\mathcal {O}}}(N_c^2)$$ O ( N c 2 ) in the quark-gluon plasma regime. In the large$$N_c$$ N c limit these regimes may become distinct phases separated by first order phase transitions. The intermediate phase has the peculiar feature that glueballs should exist and have properties that are unchanged from what is seen in the vacuum (up to$$1/N_c $$ 1 / N c corrections), while the ordinary dilute gas of mesons with broken chiral symmetry disappears and approximate chiral spin symmetry should emerge.

Physics↗

DOE OSTI2026

A Parallel Kinetic Model for Surface and Bulk Charge Storage in ε -MnO 2 Pseudocapacitors

Pseudocapacitive materials such as manganese dioxide (MnO 2 ) are attractive for energy storage applications due to their ability to combine the fast kinetics of capacitors with the higher energy density of battery-type systems. However, the electrochemical behavior of MnO 2 remains difficult to interpret mechanistically, in part because existing models often fail to distinguish between surface-based redox processes and bulk intercalation mechanisms. In this work, we develop a physics-based model that represents MnO 2 pseudocapacitance as a linear combination of two independent, parallel electrochemical processes: (i) the surface or near-surface redox storage and (ii) lithium ion intercalation into the bulk material. These two processes are treated with distinct kinetic and thermodynamic parameters and are assumed to proceed independently. The total measured current is assumed to be the sum of these two partial currents. We validate the model using rate-dependent cyclic voltammetry experiments, demonstrating that it captures key trends and provides physically interpretable parameters reflecting the relative contributions of surface and bulk processes. By enabling a clear separation between these mechanisms, the model offers a useful framework for analyzing pseudocapacitive materials and can guide the rational design of high-performance energy storage electrodes.

Energy - Storage↗

DOE OSTI2025

Disturbance and Response Model (DRM)

The DRM is a framework to interpret ecosystem process model output for 3D fire behavior model input and interpret the 3D fire behavior model output for ecosystem process model inpu

Atchley, Adam↗

DOE OSTI2025

Fusion Model for Metagenomics

This work highlights the use of an embeddings approach that can encode multiple features and create efficient contextualization of profiled metagenomes derived from microbiome samples using computer vision models and image representations of the abundance profiles. The model's embeddings can be used to cluster existing samples based on multiple conditions and interpretations, and new embeddings can be quickly created for new samples and fitted to existing clusters to characterize them. This has practical applications for unknown, unlabeled microbiome samples. The model's embeddings can be used to cluster existing samples based on multiple conditions and interpretations, and new embeddings can be quickly created for new samples and fitted to existing clusters to characterize them. This has practical applications for unknown, unlabeled microbiome samples.

Valdes, CamiloA [Lawrence Livermore National Labor↗

DOE OSTI2024

Evidential Deep Learning: Enhancing Predictive Uncertainty Estimation for Earth System Science Applications

Abstract Robust quantification of predictive uncertainty is a critical addition needed for machine learning applied to weather and climate problems to improve the understanding of what is driving prediction sensitivity. Ensembles of machine learning models provide predictive uncertainty estimates in a conceptually simple way but require multiple models for training and prediction, increasing computational cost and latency. Parametric deep learning can estimate uncertainty with one model by predicting the parameters of a probability distribution but does not account for epistemic uncertainty. Evidential deep learning, a technique that extends parametric deep learning to higher-order distributions, can account for both aleatoric and epistemic uncertainties with one model. This study compares the uncertainty derived from evidential neural networks to that obtained from ensembles. Through applications of the classification of winter precipitation type and regression of surface-layer fluxes, we show evidential deep learning models attaining predictive accuracy rivaling standard methods while robustly quantifying both sources of uncertainty. We evaluate the uncertainty in terms of how well the predictions are calibrated and how well the uncertainty correlates with prediction error. Analyses of uncertainty in the context of the inputs reveal sensitivities to underlying meteorological processes, facilitating interpretation of the models. The conceptual simplicity, interpretability, and computational efficiency of evidential neural networks make them highly extensible, offering a promising approach for reliable and practical uncertainty quantification in Earth system science modeling. To encourage broader adoption of evidential deep learning, we have developed a new Python package, Machine Integration and Learning for Earth Systems (MILES) group Generalized Uncertainty for Earth System Science (GUESS) (MILES-GUESS) ( https://github.com/ai2es/miles-guess ), that enables users to train and evaluate both evidential and ensemble deep learning. Significance Statement This study demonstrates a new technique, evidential deep learning, for robust and computationally efficient uncertainty quantification in modeling the Earth system. The method integrates probabilistic principles into deep neural networks, enabling the estimation of both aleatoric uncertainty from noisy data and epistemic uncertainty from model limitations using a single model. Our analyses reveal how decomposing these uncertainties provides valuable insights into reliability, accuracy, and model shortcomings. We show that the approach can rival standard methods in classification and regression tasks within atmospheric science while offering practical advantages such as computational efficiency. With further advances, evidential networks have the potential to enhance risk assessment and decision-making across meteorology by improving uncertainty quantification, a longstanding challenge. This work establishes a strong foundation and motivation for the broader adoption of evidential learning, where properly quantifying uncertainties is critical yet lacking.

Schreck, John S.↗

DOE OSTI2024

Improved Diagnosis of Precipitation Type with LightGBM Machine Learning

Abstract Existing precipitation-type algorithms have difficulty discerning the occurrence of freezing rain and ice pellets. These inherent biases are not only problematic in operational forecasting but also complicate the development of model-based precipitation-type climatologies. To address these issues, this paper introduces a novel light gradient-boosting machine (LightGBM)-based machine learning precipitation-type algorithm that utilizes reanalysis and surface observations. By comparing it with the Bourgouin precipitation-type algorithm as a baseline, we demonstrate that our algorithm improves the critical success index (CSI) for all examined precipitation types. Moreover, when compared with the precipitation-type diagnosis in reanalysis, our algorithm exhibits increased F1 scores for snow, freezing rain, and ice pellets. Subsequently, we utilize the algorithm to compute a freezing-rain climatology over the eastern United States. The resulting climatology pattern aligns well with observations; however, a significant mean bias is observed. We interpret this bias to be influenced by both the algorithm itself and assumptions regarding precipitation processes, which include biases associated with freezing drizzle, precipitation occurrence, and regional synoptic weather patterns. To mitigate the overall bias, we propose increasing the precipitation cutoff from 0.04 to 0.25 mm h −1 , as it better reflects the precision of precipitation observations. This adjustment yields a substantial reduction in the overall bias. Finally, given the strong performance of LightGBM in predicting mixed precipitation episodes, we anticipate that the algorithm can be effectively utilized in operational settings and for diagnosing precipitation types in climate model outputs. Significance Statement Freezing rain can have significant impacts on transportation and infrastructure, making accurate prediction of precipitation types crucial. In this study, we use a machine learning method known as LightGBM to predict precipitation types. We show that the new algorithm performs better than the existing methods for all precipitation types examined. Additionally, we compute a freezing-rain climatology over the eastern United States. Although the resulting climatology pattern corresponds well to observations, the algorithm overpredicts freezing-rain occurrence. We argue that this bias can be substantially reduced by increasing the precipitation cutoff from 0.04 to 0.25 mm h −1 . Overall, this work highlights the potential of the LightGBM algorithm for both weather forecasting and diagnosing precipitation types in climate models.

Meteorology & Atmospheric Sciences↗

DOE OSTI2024

Ground-motions site and event specificity: Insights from assessing a suite of simulated ground motions in the San Francisco Bay Area

This article presents the results of a research that is part of a larger collaborative effort between the Lawrence Berkeley National Laboratory and the Pacific Earthquake Engineering Research Center, funded by the US Department of Energy Office of Cybersecurity, Energy Security and Emergency Response. The main objective of this study is to assess a suite of near and far-field simulated ground motions obtained from 20 realizations of an M7 Hayward Fault earthquake in the San Francisco Bay Area, California USA, and inform the selection of rupture simulation parameters leading to strong motions. To this aim, comparisons are conducted with NGA-W2 and directivity ground-motion models and a selected population of records. An archetypal steel moment-resisting frame is utilized to assess infrastructure response distributions. The analyses carried out for each simulated event and subdomain with consistent properties in terms of shallow shear-wave velocity proved to be instrumental for better interpreting the differences between simulated motions and empirical models. The main reasons identified for variances between simulations and empirical relationships included (1) directivity effects fully captured by the simulations across the full breadth of rupture models; (2) site vicinity to ruptures that incorporate large-slip patches, particularly if these are in the forward-directivity direction; and (3) presence of geologic structures that can “trap” seismic waves and produce ground motions with large amplitude and long signal duration. The analyses carried out in this work provide a path for interpreting ground-motion site and event specificity obtained from a suite of physics-based simulations, differing only in the rupture model characterization, to inform the selection of simulation scenarios for site-specific engineering analyses under strong excitations. Evidence from this work points to the possibility that current hazard models may underestimate ground-motion intensities in areas where the combined effect of directivity and site conditions results in large ground-motion amplitudes.

58 GEOSCIENCES↗

DOE OSTI2025

Generalizable, fast, and accurate DeepQSPR with fastprop

Abstract Quantitative Structure–Property Relationship studies (QSPR), often referred to interchangeably as QSAR, seek to establish a mapping between molecular structure and an arbitrary target property. Historically this was done on a target-by-target basis with new descriptors being devised to specifically map to a given target. Today software packages exist that calculate thousands of these descriptors, enabling general modeling typically with classical and machine learning methods. Also present today are learned representation methods in which deep learning models generate a target-specific representation during training. The former requires less training data and offers improved speed and interpretability while the latter offers excellent generality, while the intersection of the two remains under-explored. This paper introduces , a software package and general Deep-QSPR framework that combines a cogent set of molecular descriptors with deep learning to achieve state-of-the-art performance on datasets ranging from tens to tens of thousands of molecules. provides both a user-friendly Command Line Interface and highly interoperable set of Python modules for the training and deployment of feedforward neural networks for property prediction. This approach yields improvements in speed and interpretability over existing methods while statistically equaling or exceeding their performance across most of the tested benchmarks. is designed with Research Software Engineering best practices and is free and open source, hosted at github.com/jacksonburns/fastprop.

Burns, Jackson W. (ORCID:0000000206579426)↗

DOE OSTI2026

Application of artificial intelligence methods in the international roughness index prediction of rigid and composite pavements: a systematic review

The International Roughness Index (IRI) is a widely adopted metric for quantifying pavement roughness, directly influencing vehicle safety, ride comfort, and overall roadway performance. In recent years, the use of Machine Learning (ML) models for IRI prediction has gained momentum, with the goal of improving the allocation of maintenance and rehabilitation resources by enabling accurate assessments of pavement conditions. Most prior reviews, however, have concentrated on flexible pavements, leaving a notable gap regarding rigid and composite pavements. To address this gap, the present study conducts a systematic review of Artificial Intelligence (AI) methods applied to IRI prediction for rigid and composite pavements. Literature published between 2004 and 2025 is synthesized to highlight prevailing trends, methodological contributions, and directions for future research. Particular attention is given to the types of models employed, the datasets used for training and validation, and the role of input variables and data-processing strategies. Across the included studies, ensemble learning methods (especially gradient boosting variants such as XGBoost), artificial neural networks, and hybrid architectures frequently achieved high predictive skill, with several models reporting test-set coefficients of determination approaching 0.9–0.96, indicating strong potential for capturing the influence of traffic, pavement structure, and climatic factors. Since these results are obtained from heterogeneous datasets and evaluation protocols, they are interpreted qualitatively rather than as strict cross-study rankings. Analysis of input variables revealed that pavement age and initial IRI were included in 91% (21 of 23) and 78% (18 of 23) of studies, respectively. Climatic variables such as the freezing index appeared in 57% (13 of 23), while traffic-related factors were considered in 65% (15 of 23). The findings underscore the importance of standardized, high-quality datasets, such as those from the Long-Term Pavement Performance (LTPP) program, along with data consistency, model interpretability, computational efficiency, and replicability in enhancing IRI prediction. Future research should focus on incorporating input variable selection techniques to identify the most influential predictors, thereby improving accuracy and robustness. Integrating these approaches with advanced non-linear data-driven models, coupled with robust hyperparameter optimization, holds considerable promise for strengthening the reliability of IRI prediction and supporting resilient pavement management strategies.

42 ENGINEERING↗

DOE OSTI2026

Data for FUN-PROSE: A Deep Learning Approach to Predict Condition-Specific Gene Expression in Fungi

mRNA levels of all genes in a genome is a critical piece of information defining the overall state of the cell in a given environmental condition. Being able to reconstruct such condition-specific expression in fungal genomes is particularly important to metabolically engineer these organisms to produce desired chemicals in industrially scalable conditions. Most previous deep learning approaches focused on predicting the average expression levels of a gene based on its promoter sequence, ignoring its variation across different conditions. Here we present FUN-PROSE—a deep learning model trained to predict differential expression of individual genes across various conditions using their promoter sequences and expression levels of all transcription factors. We train and test our model on three fungal species and get the correlation between predicted and observed condition-specific gene expression as high as 0.85. We then interpret our model to extract promoter sequence motifs responsible for variable expression of individual genes. We also carried out input feature importance analysis to connect individual transcription factors to their gene targets. A sizeable fraction of both sequence motifs and TF-gene interactions learned by our model agree with previously known biological information, while the rest corresponds to either novel biological facts or indirect correlations.

Genomics↗