SEARCH · Engineering Papers

Results for “Explanability”

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 487 records · Page 27

All sources NASA NTRS DOE OSTI

NASA NTRS2020

Asymmetric Surface Brightness Structure of Caustic Crossing Arc in SDSS J1226+2152: A Case for Dark Matter Substructure

We study the highly magnified arc SGAS J122651.3+215220 caused by a star-forming galaxy at z s = 2.93 crossing the lensing caustic cast by the galaxy cluster SDSS J1226+2152 (z l = 0.43), using Hubble Space Telescope observations. We report in the arc several asymmetric surface brightness features whose angular separations are a fraction of an arcsecond from the lensing critical curve and appear to be highly but unequally magnified image pairs of underlying compact sources, with one brightest pair having clear asymmetry consistently across four filters. One explanation of unequal magnification is microlensing by intracluster stars, which induces independent flux variations in the images of individual or groups of source stars in the lensed galaxy. For a second possibility, intracluster dark matter subhaloes invisible to telescopes effectively perturb lensing magnifications near the critical curve and give rise to persistently unequal image pairs. Our modelling suggests, at least for the most prominent identified image pair, that the microlensing hypothesis is in tension with the absence of notable asymmetry variation over a six-year baseline, while subhaloes of ∼10 6 –10 8 \Mꙩ anticipated from structure formation with cold dark matter typically produce stationary and sizable asymmetries. We judge that observations at additional times and more precise lens models are necessary to stringently constrain temporal variability and robustly distinguish between the two explanations. The arc under this study is a scheduled target of a Director’s Discretionary Early Release Science program of the James Webb Space Telescope, which will provide deep images and a high-resolution view with integral field spectroscopy.

Liang Dai↗

NASA NTRSDate not supplied

Explaining Soft-Goal Conflicts through Constraint Relaxations

Recent work suggests to explain trade-offs between soft goals in terms of their conflicts, i. e., minimal unsolvable soft-goal subsets. But this does not explain the conflicts themselves: Why can a given set of soft-goals not be jointly achieved? Here we approach that question in terms of the underlying constraints on plans in the task at hand, namely resource availability and time windows. In this context, a natural form of explanation for a soft-goal conflict is a minimal constraint relaxation under which the conflict disappears (“if the deadline was 1 hour later, it would work”). We explore algorithms for computing such explanations. A baseline is to simply loop over all relaxed tasks and compute the conflicts for each separately. We improve over this by two algorithms that leverage information – conflicts, reachable states – across relaxed tasks. We show that these algorithms can exponentially outperform the baseline in theory, and we run experiments confirming that advantage in practice.

Planning↗

NASA NTRSDate not supplied

Large Impact Features on Icy Galilean Satellites: Implications for Ice Shell Thickness and Habitability.

Nowhere in the solar system are impact morphologies observed in greater variety than on the icy Galilean satellites. This is very likely a consequence of the structural and thermal state of the crust at the time of impact, and perhaps impact velocity. We have conducted a multi-disciplinary investigation to study how these features formed. The particular combination of geophysical factors and impactor characteristics that is shared by these satellites is likely responsible for these features. We have derived Digital Terrain Models for a number of these features, which have been used as a tool for producing facies maps. Impact features on Ganymede and Callisto are relatable to impact features on Europa, with implications for its ice shell history. We have identified two broad classes of these impact features: (a) Crater forms, which variously include pits, central domes, and crater rims; and (b) Palimpsests (including pene-palimpsests), which all have smooth enclosed central plains, an extensive outer undulating plains unit, concentric arcuate ridges(in the case of pene-palimpsests), and no recognizable rim. The very large, measured volume of the undulating plains unit of the pene-palimpsest Buto, Ganymede, substantially exceeds the amount of target material that could have been both melted and excavated by the impact event itself. Therefore, we conclude that the undulating plains material was derived from a fluid or “slushy” layer in the near-surface target at the time of impact. The recognition of undulating plains as the dominant unit in Memphis and Nidaba, two other palimpsests on Ganymede that we have mapped(which display highly variable crater counts, with Buto being the least cratered),leads us to a working hypothesis that these plains only form where and when the target has substantial near surface fluid or “slush.” By extension, all other impact features on Ganymede and Callisto, such as those with central pits and central domes, formed in targets with no pre-existing, or at most inconsequential, amounts of fluid or “slush.” Supporting this conclusion, modeling by members of our group indicates that the present topography of impact features with central pits and domes can be largely explained by the behavior of ice alone, with no or little contribution from fluids. Our study supports explanations for the formation and final appearance of palimpsests (and pene-palimpsests) that are entirely applicable to this feature class, and sharply distinct from explanations for other impact features (crater forms) on Ganymede and Callisto.

Jeffrey M Moore↗

NASA NTRSDate not supplied

On the Solar Open Magnetic Flux Problem

The solar magnetic fields emerging from the photosphere into the chromosphere and corona are comprised of a combination of “closed” (field lines with both ends rooted at the Sun) and “open” (field lines with only one end at the Sun) fields. Since the early 2000’s, the magnitude of total unsigned open magnetic flux estimated by coronal models has been in significant disagreement with in situ spacecraft observations, especially during solar maximum. Estimates of total open unsigned magnetic flux using coronal hole observations (e.g., using extreme ultraviolet (EUV) or Helium (He) are in general agreement with the coronal model results and thus show similar disagreements with in situ observations. This paper provides a brief overview of the problem, summarizes the proposed explanations for the discrepancies, and presents results that strongly support the explanation that the discrepancy is due to dynamics at the open-closed boundary. These results are derived from determination of the total unsigned open magnetic flux, utilizing the Wang-Sheeley-Arge (WSA) model at a particular spatial resolution and different field-line tracing methods. One of these methods produces excellent agreement with in situ observations. Our results imply that active regions residing near the boundaries of mid-latitude coronal holes are the primary source of the missing open flux. Furthermore, they resolve many of the seemingly contradictory facts that have made resolving the open-flux problem so difficult.

Charles Arge↗

NASA NTRS2024

Ultra-deep ATCA imaging of 47 Tucanae reveals a central compact radio source

We present the results of an ultradeep radio continuum survey, containing ∼480 hr of observations, of the Galactic globular cluster 47 Tucanae with the Australia Telescope Compact Array. This comprehensive coverage of the cluster allows us to reach rms noise levels of 1.19 μJy beam −1 at 5.5 GHz, 940 nJy beam−1 at 9 GHz, and 790 nJy beam −1 in a stacked 7.25 GHz image. This is the deepest radio image of a globular cluster and the deepest image ever made with the Australia Telescope Compact Array. We identify ATCA J002405.702-720452.361, a faint (6.3 ± 1.2 μJy at 5.5 GHz, 5.4 ± 0.9 μJy at 9 GHz), flat-spectrum (α = −0.31 ± 0.54) radio source that is positionally coincident with the cluster center and potentially associated with a faint X-ray source. No convincing optical counterpart was identified. We use radio, X-ray, optical, and UV data to show that explanations involving a background active galactic nucleus, a chromospherically active binary, or a binary involving a white dwarf are unlikely. The most plausible explanations are that the source is an undiscovered millisecond pulsar or a weakly accreting black hole. If the X-ray source is associated with the radio source, the fundamental plane of black-hole activity suggests a black hole mass of ∼54–6000 M ⊙ , indicating an intermediate-mass black hole or a heavy stellar-mass black hole.

Alessandro Paduano↗

NASA NTRS2024

Volatiles in melt inclusions from lunar mare basalts: Bridging the gap in the H 2 O/Ce ratio between melt Inclusions in lunar pyroclastic sample 74220 and other mare samples

The H2O concentration and H 2 O/Ce ratio in olivine-hosted melt inclusions are high in lunar pyroclastic sample 74220 (H 2 O up to 1410 ppmw; H 2 O/Ce up to 77) but lower (H2O 10 to 430 ppmw; H 2 O/Ce 0.3 to 9.4) in all other lunar samples studied before this work. The difference in H 2 O concentration and in H 2 O/Ce ratio is absent for other volatile elements (F, S, and Cl) in melt inclusions in 74220 and other lunar samples. Because H 2 O (or H) is a critical volatile component with significant ramifications on the origin and evolution of the Moon, it is important to understand what causes such a large gap in H 2 O/Ce ratio between 74220 and other lunar samples. Two explanations have been advanced. One is that volcanic product in sample 74220 has the highest cooling rate and thus best preserved H 2 O in melt inclusions compared to melt inclusions in other samples. The other explanation is that sample 74220 comes from a localized heterogeneity enriched in some volatiles. To distinguish these two possibilities, here we present new data from two rapidly cooled lunar samples with glassy melt inclusions: olivine-hosted melt inclusions (OHMIs) in 79135 regolith breccia (unknown cooling rate but with glassy MIs similar in texture with those in 74220), and pyroxene-hosted melt inclusions (PHMIs) in 15597 pigeonite basalts (known high cooling rate, second only to 74220 and 15421). In addition, we also investigated new OHMIs in sample 74220. If the gap is due to the difference in cooling rates, samples with cooling rates between those of 74220 and other studied lunar samples should have preserved intermediate H2O concentrations and H 2 O/Ce ratios. Our results show that melt inclusions in 79135 and 15597 contain high H 2 O concentrations (up to 969 ppmw in 79135 and up to 793 ppmw in 15597) and high H 2 O/Ce ratios (up to 21 in 79135 and up to 13 in 15997), bridging the big gap in H 2 O/Ce ratio among 74220 and other lunar samples. Combined with literature data, we confirm that H 2 O/Ce ratios of different lunar samples are positively correlated to the cooling rates and independent of the type of mare basalts. We hence reinforce the interpretation that the lunar sample with the highest cooling rate best represents pre-eruptive volatiles in lunar basalts due to the least degassing. Based on Ce concentration in the primitive lunar mantle, we estimate that H 2 O concentration in the primitive lunar mantle (meaning bulk silicate Moon) is 121 ± 15 ppmw. Our new data also further constrain F/P, S/Dy and Cl/Ba ratios in lunar basalts and the lunar mantle. Estimated F, P, and S concentrations in the lunar primitive mantle are 4.4 ± 1.1 ppmw, 22 ± 8 ppmw, and 67 +67 −33 ppmw, respectively.

Xue Su↗

NASA NTRS2026

An Investigation Regarding the Effects of Volume Dynamics on Gas Turbine Engine Performance and Operability

This work explores the impact of adding volume dynamics to an electrified turbofan engine model. The zero-dimensional single-input single-output volume elements capture the physics of fluid compressibility of flow in the engine core gas path turbomachinery during transient operation with greater fidelity. Some background regarding the motivation for this effort is provided. A model of the volume element is developed from first principles. The baseline engine model is described and an explanation is given of the modifications made to incorporate the volume elements into the new model. Details of the simulated inputs, designed to elicit exaggerated dynamic responses and to facilitate comparison between the baseline and modified engine models, are explicated. The results of these simulations are presented and discussed, demonstrating that the transient performance and operability of the modified engine model, measured by parameters such as shaft speeds, thrust, and high-pressure compressor (HPC) stall margin, are impacted by the additional dynamics. Commentary is provided regarding potential explanations for the differences in behavior exhibited by the two models. Finally, suggestions are offered for future work that might improve the volume models developed in this effort and leverage the improved fidelity enabled by the modeling of turbomachinery volume dynamics.

Volume dynamics↗

DOE OSTI2024

Dark Photon Search at the Short-Baseline Near Detector

Neutrino physics has long been a key field in elementary particle physics, both enhancing our understanding of the Standard Model (SM) and raising new questions. Among these are the so-called "Short-Baseline Anomalies" observed by neutrino experiments, particularly the MiniBooNE experiment at Fermilab, which detected an excess of low-energy electron-like events. In recent years, beyond Standard Model (BSM) explanations have been proposed to address this anomaly, with a focus on neutrino beam-related processes. A novel interpretation involving a dark-sector explanation was recently suggested, introducing a vector portal that connects the SM and dark sectors through a new interaction mediated by a bosonic particle, the Dark Photon. This work investigates the production of dark-sector particles, specifically Dark Photons, in the Booster Neutrino Beam (BNB) at Fermilab and their potential detection at the Short-Baseline Near Detector (SBND). The BNB produces mesons which decay into Dark Photons, detectable via their decay into electron-positron pairs at SBND. By exploiting the temporal structure of the neutrino beam, we propose a method to isolate Dark Photon signals from neutrino backgrounds using time-delayed event detection. In this thesis, the sensitivity of SBND to Dark Photons is assessed using a three-year exposure, demonstrating that SBND has the potential to significantly improve current experimental constraints on Dark Photons. This analysis provides a promising avenue for future dark sector searches in neutrino experiments.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Effects of spatial variability in vegetation phenology, climate, landcover, biodiversity, topography, and soil property on soil respiration across a coastal ecosystem

Coastal terrestrial-aquatic interfaces (TAIs) are crucial contributors to global biogeochemical cycles and carbon exchange. A systematic evaluation of the interaction between coastal catchment properties and carbon dioxide (CO2) emission by soil respiration is significant for assessing carbon dynamics and predicting the future trajectory of atmospheric CO2 concentrations in coastal TAIs. The soil CO2 efflux in these transition zones is however poorly understood due to the high spatiotemporal dynamics of TAIs, as various sub-ecosystems in this region are compressed and expanded by complex influences of tides, changes in river levels, climate, and land use. We focus on the Chesapeake Bay region to (i) investigate the spatial heterogeneity of the coastal ecosystem and identify spatial zones with similar environmental characteristics based on the spatial data layers, including vegetation index (kNDVI), climate, landcover, diversity, topography, soil property, and relative tidal elevation; (ii) understand the primary driving factors affecting soil respiration within sub-ecosystems of the coastal ecosystem. Specifically, we employed hierarchical clustering analysis to identify spatial regions with distinct environmental characteristics, followed by the determination of main driving factors using Random Forest regression and SHapley Additive exPlanations. Maximum and minimum temperature are the main drivers common to all sub-ecosystems, while each region also has additional unique major drivers that differentiate them from one another. Precipitation exerts an influence on vegetated lands, while soil pH value holds importance specifically in forested lands. In croplands characterized by high clay content and low sand content, the significant role is attributed to bulk density. Wetlands demonstrate the importance of both elevation and sand content, with clay content being more relevant in non-inundated wetlands than in inundated wetlands. The topographic wetness index significantly contributes to the mixed vegetation areas, including shrub, grass, pasture, and forest. Additionally, our research reveals that dense vegetation land covers and urban/developed areas exhibit distinct soil property drivers. Overall, there is no one-size-fits-all approach to modeling carbon fluxes in coastal TAIs, and our study highlights the importance of further research and monitoring practices to improve our understanding of carbon dynamics and promote the sustainable management of coastal TAIs.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2021

Deep Learning Estimation of Daily Ground–Level NO 2 Concentrations from Remote Sensing Data

The limited number of nitrogen dioxide (NO 2 ) surface measurements calls for the development of highly accurate approaches to estimating surface NO 2 concentrations. In this study, we leverage a new satellite instrument, the TROPOspheric Monitoring Instrument (TROPOMI), along with other predictor variables, to estimate daily surface NO 2 concentrations over Texas in 2019. We use the deep convolutional neural network (Deep-CNN), an advanced deep learning algorithm, to obtain estimates and achieve a correlation coefficient (R) of 0.91, an index of agreement (IOA) of 0.95, and a mean absolute bias (MAB) of 1.75 ppb in surface NO 2 estimation. Additionally, we leverage a novel approach, SHapley Additive exPlanations (SHAP), to describe how Deep-CNN understands each predictor variable. The SHAP results show that the Deep-CNN model has an advanced understanding of the dataset, revealing that TROPOMI closely captures levels of NO 2 . In addition, we show the superiority of our Deep-CNN model at estimating surface NO 2 over other well-known machine learning and regression models in the field, including the support vector machines (SVM), random forest (RF), and multiple linear regression (MLR). Although SVM and RF show strong capabilities at estimating surface NO 2 concentrations, their accuracy is inferior to that of the Deep-CNN model, ranking second and third in model accuracy in this study. The MLR, however, shows a poor ability at NO 2 estimation and ranks last among all models. Furthermore, testing the impact of sample size on model performance, we also show that, compared to other models, Deep-CNN needs more samples to trigger its strength at surface NO 2 estimation.

54 ENVIRONMENTAL SCIENCES↗

DOE OSTI2026

Contrasting Carbon–Water–Energy Dynamics in Perennial and Annual Bioenergy Agroecosystems Using Eddy Covariance and Interpretable Machine Learning

Understanding how agroecosystems respond to environmental variability is fundamental to predicting productivity and sustainability under a changing climate. We analyzed 55 site-years of high-frequency eddy covariance observations from five agroecosystems—two perennial grasses (miscanthus and switchgrass), two annual rotation systems (maize–soybean and sorghum–soybean), and a restored native prairie—to examine ecosystem-scale carbon, water, and energy fluxes. Using an interpretable machine-learning framework with regression tree ensembles, Shapley Additive Explanations, and Accumulated Local Effects, we quantified how environmental and temporal factors regulate gross primary productivity (GPP), evapotranspiration (ET), water-use efficiency, and the Bowen ratio. Perennials exhibited stronger physiological buffering and maintained fluxes across a broader range of temperature and moisture conditions, reflecting deeper rooting and persistent canopy cover. Annuals, in contrast, showed greater short-term variability and stronger coupling to atmospheric demand, with GPP and ET declining rapidly under low humidity or soil moisture. Differences in temperature sensitivity of Bowen ratio further revealed that perennials sustained proportionally greater sensible heat flux under cool conditions, whereas annuals exhibited constrained energy exchange when evaporative demand was low. Together, these results demonstrate that crop life cycle and canopy structure are fundamental determinants of ecosystem-scale carbon–water–energy coupling. By integrating long-term flux observations with interpretable machine learning, this study identifies the environmental drivers that shape agroecosystem function and highlights how conversion from annual to perennial feedstocks can enhance climatic resilience and alter land–atmosphere energy feedbacks. These findings provide a data-driven basis for improving crop and Earth-system models and for guiding bioenergy landscape design under future climate scenarios.

Accumulated Local Effects↗

DOE OSTI2026

Data, model inputs, and analysis scripts associated with a manuscript on stream intermittency controls across spatial scales in Pacific Northwest watersheds

NOTE: The manuscript associated with this data package is currently in review. The data may be revised based on reviewer feedback. Upon manuscript acceptance, this data package will be updated with the final dataset and additional metadata. This data package is associated with the manuscript "Hydroclimatic Memory and Watershed Template Shape Stream Intermittency: Multi-scale Attribution Using Process-based Simulation and Explainable ML" by Niroula et al. (2026), submitted to Water Resources Research (WRR). The study investigates the dominant controls on stream intermittency across local, reach, and watershed scales using a coupled process-based simulation and explainable machine-learning framework. Long-term daily simulations from the Advanced Terrestrial Simulator (ATS) were used to generate wetness states and ponded-depth responses over river-corridor cells. These ATS outputs were then aggregated across scales and used to train XGBoost (eXtreme Gradient Boosting) models. SHAP (SHapley Additive exPlanations) was applied to quantify the relative importance of hydroclimatic forcings, watershed template attributes, and antecedent-memory effects in shaping intermittency behavior. The analysis is carried out for three contrasting Pacific Northwest watersheds: Oak Creek (OCW), American River Watershed (ARW), and H.J. Andrews (HJA). Across these testbeds, the package contains ATS-ready watershed inputs, ATS run configuration and selected output files, model-evaluation data products, intermittency-analysis datasets, machine-learning target-feature tables, SHAP outputs, and notebooks used to organize, analyze, and visualize results. At a high level, the package documents a workflow in which ATS provides the physically based simulation backbone and explainable machine learning is used as a post-processing attribution tool. The contents are intended to support interpretation of the manuscript figures and results, provide context for how intermittency metrics were generated at multiple scales, and preserve the key artifacts needed to understand and reuse the analysis workflow. The package contains a high-level directory summary file (`summary.txt`) and four main content folders (1) `evaluation_plots` contains evaluation figures and supporting evaluation datasets; (2) `intermittency_plots` contains intermittency-focused analysis notebook and prepared datasets; (3) `ml-training-and-shap_values_plots` contains ML training inputs, SHAP outputs, and figure-generation notebooks; and (4) `watershed_mesh_and_ats_input` contains ATS model setup materials, forcing inputs, geometry, and selected run files. More specifically, the `evaluation_plots` folder contains the notebook used for ATS evaluation plotting and site-specific evaluation datasets. These include evapotranspiration and water-balance products for three watersheds, as well as an Oak Creek field-measurement discharge file. The `intermittency_plots` folder contains the notebook used for intermittency analysis and the prepared datasets used to analyze intermittent and non-intermittent wetness behavior across the study watersheds. The `ml-training-and-shap_values_plots` folder contains notebooks and outputs for the machine-learning and explainability workflow. This includes the main XGBoost and SHAP notebook(s), a beeswarm plotting notebook, target-feature tables for machine-learning training, SHAP summary tables, and per-sample SHAP value archives. The `watershed_mesh_and_ats_input` folder contains ATS-related watershed inputs and supporting materials. This includes mesh and shape products, ATS-readable LAI and meteorological forcing inputs, selected ATS spinup and transient-run files, and a watershed workflow example notebook. Subdirectories are organized by watershed where applicable.All files are .cpg (codepage files), .csv (comma-separated values), .dbf (database files), .exo (Exodus mesh format), .h5 (HDF5 format), .ipynb (Jupyter notebooks), .pkl (Python pickle), .prj (projection files), .sh (shell scripts), .shp (shapefile geometry), .shx (shapefile index), .txt (text files), or .xml (markup data).

Advanced Terrestrial Simulator↗

DOE OSTI2026

Data and scripts associated with a manuscript analyzing ELM-FATES parameter sensitivity under pre-fire and postfire scenarios using machine learning

Aboveground biomass↗

DOE OSTI2022

Identifying Critical Infrastructure in Imagery Data Using Explainable Convolutional Neural Networks

To date, no method utilizing satellite imagery exists for detailing the locations and functions of critical infrastructure across the United States, making response to natural disasters and other events challenging due to complex infrastructural interdependencies. This paper presents a repeatable, transferable, and explainable method for critical infrastructure analysis and implementation of a robust model for critical infrastructure detection in satellite imagery. This model consists of a DenseNet-161 convolutional neural network, pretrained with the ImageNet database. The model was provided additional training with a custom dataset, containing nine infrastructure classes. The resultant analysis achieved an overall accuracy of 90%, with the highest accuracy for airports (97%), hydroelectric dams (96%), solar farms (94%), substations (91%), potable water tanks (93%), and hospitals (93%). Critical infrastructure types with relatively low accuracy are likely influenced by data commonality between similar infrastructure components for petroleum terminals (86%), water treatment plants (78%), and natural gas generation (78%). Local interpretable model-agnostic explanations (LIME) was integrated into the overall modeling pipeline to establish trust for users in critical infrastructure applications. The results demonstrate the effectiveness of a convolutional neural network approach for critical infrastructure identification, with higher than 90% accuracy in identifying six of the critical infrastructure facility types.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Development of Short-Term Forecasting Models Using Plant Asset Data and Feature Selection

Nuclear power plants collect and store large volumes of heterogeneous data from various components and systems. With recent advances in machine learning (ML) techniques, these data can be leveraged to develop diagnostic and short-term forecasting models to better predict future equipment condition. Maintenance operations can then be planned in advance whenever degraded performance is predicted, thus resulting in fewer unplanned outages and the optimization of maintenance activities. This enables lower maintenance costs and improves the overall economics of nuclear power. This paper focuses on developing a short-term forecasting process that leverages a feature selection process to distill large volumes of heterogeneous data and predict specific equipment parameters. A variety of feature selection methods, including Shapley Additive Explanations (SHAP) and variance inflation factor (VIF), were used to select the optimal features as inputs for three ML methods: long short-term memory (LSTM) networks, support vector regression (SVR), and random forest (RF). Each combination of model and input features was used to predict a pump bearing temperature both 1 and 24 hours in advance, based on actual plant system data. The optimal inputs for the LSTM and SVR were selected using the SHAP values, while the optimal input for the RF consisted solely of the response variable itself. Each model produced similar 1-hour-ahead predictions, with root mean square errors (RMSEs) of roughly 0.006. For the 24-hour-ahead predictions, differences could be seen between LSTM, SVR, and RF, as reflected by model performances of 0.036 +- 0.014, 0.0026 +- 0, and 0.063 +- 0.004 RMSE, respectively. As big data and continuous online monitoring become more widely available, the proposed feature selection process can be used for many applications beyond the prediction of process parameters within nuclear infrastructure.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

NASA NTRSDate not supplied

Understanding and Verifying Neural Networks

Deep Neural Networks (DNNs) have gained immense popularity in recent times and have widespread use in applications such as image classification, sentiment analysis, speech recognition and also in safety-critical applications such as autonomous driving. However, they suffer limitations such as lack of explainability and robustness which raise safety and security concerns in their usage. Further, the complex structure and large input spaces of DNNs act as an impediment to thorough verification and testing. The SafeDNN project at the Robust Software Engineering (RSE) group at NASA aims at exploring techniques to ensure that systems that use deep neural networks are safe, robust and interpretable. In this talk, I will be presenting our technique Prophecy that automatically infers formal properties of deep neural network models. The tool extracts patterns based on neuron activations as preconditions that imply certain desirable output properties of the model. I would be highlighting case studies that use Prophecy in obtaining explanations for network decisions, understanding correct and incorrect behavior, providing formal guarantees wrt safety and robustness, and debugging neural network models. We have applied the tool on image classification networks, neural network controllers providing turn advisories in unmanned aircrafts, regression models used for autonomous center-line tracking in aircrafts and neural network object detectors

Deep Neural Networks↗

NASA NTRS2022

Using an Explainable Machine Learning Approach to Characterize Earth System Model Errors: Application of SHAP Analysis to Modeling Lightning Flash Occurrence

Computational models of the Earth System are critical tools for modern scientific inquiry. Effortstoward evaluating and improving errors in representations of physical and chemical processes inthese large computational systems are commonly stymied by highly nonlinear and complexerror behavior. Recent work has shown that these errors can be effectively predicted usingmodern Artificial Intelligence (A.I.) techniques. In this work, we go beyond these previousstudies to apply an interpretable A.I. technique to not only predict model errors but also movetoward understanding the underlying reasons for successful error prediction. We use XGBoostclassification trees and SHapley Additive exPlanations (SHAP) analysis to explore the errors inthe prediction of lightning occurrence in the NASA GEOS model, a widely used Earth SystemModel. This explainable error prediction system can effectively predict the model error andindicates that the errors are strongly related to convective processes and the characteristics ofthe land surface.

Artificial intelligence↗

DOE OSTI2022

The first crystal structure of a xylobiose-bound xylobiohydrolase with high functional specificity from the bacterial glycoside hydrolase family 30, subfamily 10

Xylobiose is a prebiotic sugar that has applications in functional foods. This report describes the first X-ray crystallographic structure models of apo and xylobiose-bound forms of a xylobiohydrolase (XBH) from Acetivibrio clariflavus. This xylan-active enzyme, a member of the recently described glycoside hydrolase family 30 (GH30), subfamily 10, phylogenetic clade has been shown to strictly release xylobiose as its primary hydrolysis product. Inspection of the apo structure reveals a glycone region X-2-binding slot. When X-2 binds, the non-reducing xylose in the -2 subsite is highly coordinated with numerous hydrogen bond contacts while contacts in the -1 subsite mostly reflect interactions typical for GH30 and enzymes in clan A of the carbohydrate-active enzymes database (CAZy). This structure provides an explanation for the high functional specificity of this new bacterial GH30 XBH subfamily.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗