Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “pattern classification”

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 631 records · Page 35
All sourcesNASA NTRSDOE OSTI
NASA NTRS1992

Evolution of EF-hand calcium-modulated proteins. II. Domains of several subfamilies have diverse evolutionary histories

In the first report in this series we described the relationships and evolution of 152 individual proteins of the EF-hand subfamilies. Here we add 66 additional proteins and define eight (CDC, TPNV, CLNB, LPS, DGK, 1F8, VIS, TCBP) new subfamilies and seven (CAL, SQUD, CDPK, EFH5, TPP, LAV, CRGP) new unique proteins, which we assume represent new subfamilies. The main focus of this study is the classification of individual EF-hand domains. Five subfamilies--calmodulin, troponin C, essential light chain, regulatory light chain, CDC31/caltractin--and three uniques--call, squidulin, and calcium-dependent protein kinase--are congruent in that all evolved from a common four-domain precursor. In contrast calpain and sarcoplasmic calcium-binding protein (SARC) each evolved from its own one-domain precursor. The remaining 19 subfamilies and uniques appear to have evolved by translocation and splicing of genes encoding the EF-hand domains that were precursors to the congruent eight and to calpain and to SARC. The rates of evolution of the EF-hand domains are slower following formation of the subfamilies and establishment of their functions. Subfamilies are not readily classified by patterns of calcium coordination, interdomain linker stability, and glycine and proline distribution. There are many homoplasies indicating that similar variants of the EF-hand evolved by independent pathways.

Non-NASA Center↗
NASA NTRS2017

Daily Kilometer-Scale MODIS Satellite Maps of PM2.5 Describe Wintertime Episodes

The San Joaquin Valley (SJV) suffers from severe health-endangering episodes of PM2.5 aerosol loadings in wintertime; episodes last approximately 5 days and differ in geographical distribution and composition. PM2.5 stations are scattered; consequently the use of remote sensing to map variable regional patterns of these varying respirable aerosol concentrations is desirable. High-precision AOT retrievals can capture column particulate loading. However,PM2.5 mapping is challenging due to several reasons: particularly thin mixed layers (ML) and thus relatively low aerosol optical thickness (AOT) close to current measurement limits, variable and a typical composition of the aerosols, and complex surface bidirectional reflectance. However, the West does present some advantages in analysis. Air basins are isolated from long-distance transport, and experience predominant strong meteorological subsidence. Thus these Western basin regions have fewer problematic cases of overriding aerosol layers detached from the surface. To counter such local overriding, Chu et al. have described an approach for the Eastern US, and He et al have described a synoptic classification approach useful in Shanghai. The Bay Area Air Quality Management District (BAAQMD) expands our experience with the use of AOT, with lower PM2.5 and several isolated sub-basins. We have prepared daily maps of episodes in each region. We present also a sequence of increasingly detailed statistical models, AOT initially appears to contribute little information; however, inclusion of weather information reveals its utility. Lyapustin and Wang's MultiAngle Implementation of Atmospheric Correction (MAIAC) retrieval for AOT provided the most useful operational remote sensing information for these regions. It provides high (1-km) spatial resolution maps and a high percentage of availability. Empirical regression methods have found that random effects regression models (aka mixed effects models, ME) employing AOT provide good estimates of ground PM2.5 concentrations.Here, we attempt to extend these methods and evaluate the usefulness of AOT with greater physical analysis, based on DISCOVER-AQ4 experience.

Chatfield, Robert B.↗
DOE OSTI2023

KG-Hub—building and exchanging biological knowledge graphs

Knowledge graphs (KGs) are a powerful approach for integrating heterogeneous data and making inferences in biology and many other domains, but a coherent solution for constructing, exchanging, and facilitating the downstream use of KGs is lacking. Here we present KG-Hub, a platform that enables standardized construction, exchange, and reuse of KGs. Features include a simple, modular extract–transform–load pattern for producing graphs compliant with Biolink Model (a high-level data model for standardizing biological data), easy integration of any OBO (Open Biological and Biomedical Ontologies) ontology, cached downloads of upstream data sources, versioned and automatically updated builds with stable URLs, web-browsable storage of KG artifacts on cloud infrastructure, and easy reuse of transformed subgraphs across projects. Current KG-Hub projects span use cases including COVID-19 research, drug repurposing, microbial–environmental interactions, and rare disease research. KG-Hub is equipped with tooling to easily analyze and manipulate KGs. KG-Hub is also tightly integrated with graph machine learning (ML) tools which allow automated graph ML, including node embeddings and training of models for link prediction and node classification.

59 BASIC BIOLOGICAL SCIENCES↗
NASA NTRS1988

Cluster analysis of multiple planetary flow regimes

A modified cluster analysis method developed for the classification of quasi-stationary events into a few planetary flow regimes and for the examination of transitions between these regimes is described. The method was applied first to a simple deterministic model and then to a 500-mbar data set for Northern Hemisphere (NH), for which cluster analysis was carried out in the subspace of the first seven empirical orthogonal functions (EOFs). Stationary clusters were found in the low-frequency band of more than 10 days, while transient clusters were found in the band-pass frequency window between 2.5 and 6 days. In the low-frequency band, three pairs of clusters determined EOFs 1, 2, and 3, respectively; they exhibited well-known regional features, such as blocking, the Pacific/North American pattern, and wave trains. Both model and low-pass data exhibited strong bimodality.

Mo, Kingtse↗
NASA NTRS1995

Accurate estimation of sigma(exp 0) using AIRSAR data

During recent years signature analysis, classification, and modeling of Synthetic Aperture Radar (SAR) data as well as estimation of geophysical parameters from SAR data have received a great deal of interest. An important requirement for the quantitative use of SAR data is the accurate estimation of the backscattering coefficient sigma(exp 0). In terrain with relief variations radar signals are distorted due to the projection of the scene topography into the slant range-Doppler plane. The effect of these variations is to change the physical size of the scattering area, leading to errors in the radar backscatter values and incidence angle. For this reason the local incidence angle, derived from sensor position and Digital Elevation Model (DEM) data must always be considered. Especially in the airborne case, the antenna gain pattern can be an additional source of radiometric error, because the radar look angle is not known precisely as a result of the the aircraft motions and the local surface topography. Consequently, radiometric distortions due to the antenna gain pattern must also be corrected for each resolution cell, by taking into account aircraft displacements (position and attitude) and position of the backscatter element, defined by the DEM data. In this paper, a method to derive an accurate estimation of the backscattering coefficient using NASA/JPL AIRSAR data is presented. The results are evaluated in terms of geometric accuracy, radiometric variations of sigma(exp 0), and precision of the estimated forest biomass.

Holecz, Francesco↗
DOE OSTI2022

The miniJPAS survey quasar selection – I. Mock catalogues for classification

In this series of papers, we employ several machine learning (ML) methods to classify the point-like sources from the miniJPAS catalogue, and identify quasar candidates. Since no representative sample of spectroscopically confirmed sources exists at present to train these ML algorithms, we rely on mock catalogues. In this first paper, we develop a pipeline to compute synthetic photometry of quasars, galaxies, and stars using spectra of objects targeted as quasars in the Sloan Digital Sky Survey . To match the same depths and signal-to-noise ratio distributions in all bands expected for miniJPAS point sources in the range 17.5 ≤ r < 24, we augment our sample of available spectra by shifting the original r-band magnitude distributions towards the faint end, ensure that the relative incidence rates of the different objects are distributed according to their respective luminosity functions, and perform a thorough modelling of the noise distribution in each filter, by sampling the flux variance either from Gaussian realizations with given widths, or from combinations of Gaussian functions. Finally, we also add in the mocks the patterns of non-detections which are present in all real observations. Although the mock catalogues presented in this work are a first step towards simulated data sets that match the properties of the miniJPAS observations, these mocks can be adapted to serve the purposes of other photometric surveys.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRS2009

Quantifying Parameter Sensitivity, Interaction and Transferability in Hydrologically Enhanced Versions of Noah-LSM over Transition Zones

We use sensitivity analysis to identify the parameters that are most responsible for shaping land surface model (LSM) simulations and to understand the complex interactions in three versions of the Noah LSM: the standard version (STD), a version enhanced with a simple groundwater module (GW), and version augmented by a dynamic phenology module (DV). We use warm season, high-frequency, near-surface states and turbulent fluxes collected over nine sites in the US Southern Great Plains. We quantify changes in the pattern of sensitive parameters, the amount and nature of the interaction between parameters, and the covariance structure of the distribution of behavioral parameter sets. Using Sobol s total and first-order sensitivity indexes, we show that very few parameters directly control the variance of the model output. Significant parameter interaction occurs so that not only the optimal parameter values differ between models, but the relationships between parameters change. GW decreases parameter interaction and appears to improve model realism, especially at wetter sites. DV increases parameter interaction and decreases identifiability, implying it is overparameterized and/or underconstrained. A case study at a wet site shows GW has two functional modes: one that mimics STD and a second in which GW improves model function by decoupling direct evaporation and baseflow. Unsupervised classification of the posterior distributions of behavioral parameter sets cannot group similar sites based solely on soil or vegetation type, helping to explain why transferability between sites and models is not straightforward. This evidence suggests a priori assignment of parameters should also consider climatic differences.

Rosero, Enrique↗
DOE OSTI2024

Preserving isohydricity: vertical environmental variability explains Amazon forest water-use strategies

Abstract Increases in hydrological extremes, including drought, are expected for Amazon forests. A fundamental challenge for predicting forest responses lies in identifying ecological strategies which underlie such responses. Characterization of species-specific hydraulic strategies for regulating water-use, thought to be arrayed along an ‘isohydric–anisohydric’ spectrum, is a widely used approach. However, recent studies have questioned the usefulness of this classification scheme, because its metrics are strongly influenced by environments, and hence can lead to divergent classifications even within the same species. Here, we propose an alternative approach positing that individual hydraulic regulation strategies emerge from the interaction of environments with traits. Specifically, we hypothesize that the vertical forest profile represents a key gradient in drought-related environments (atmospheric vapor pressure deficit, soil water availability) that drives divergent tree water-use strategies for coordinated regulation of stomatal conductance (gs) and leaf water potentials (ΨL) with tree rooting depth, a proxy for water availability. Testing this hypothesis in a seasonal eastern Amazon forest in Brazil, we found that hydraulic strategies indeed depend on height-associated environments. Upper canopy trees, experiencing high vapor pressure deficit (VPD), but stable soil water access through deep rooting, exhibited isohydric strategies, defined by little seasonal change in the diurnal pattern of gs and steady seasonal minimum ΨL. In contrast, understory trees, exposed to less variable VPD but highly variable soil water availability, exhibited anisohydric strategies, with fluctuations in diurnal gs that increased in the dry season along with increasing variation in ΨL. Our finding that canopy height structures the coordination between drought-related environmental stressors and hydraulic traits provides a basis for preserving the applicability of the isohydric-to-anisohydric spectrum, which we show here may consistently emerge from environmental context. Our work highlights the importance of understanding how environmental heterogeneity structures forest responses to climate change, providing a mechanistic basis for improving models of tropical ecosystems.

Forestry↗
DOE OSTI2020

Metabolic Response in Patients With Post-treatment Lyme Disease Symptoms/Syndrome

Abstract Background Post-treatment Lyme disease symptoms/syndrome (PTLDS) occurs in approximately 10% of patients with Lyme disease following antibiotic treatment. Biomarkers or specific clinical symptoms to identify patients with PTLDS do not currently exist and the PTLDS classification is based on the report of persistent, subjective symptoms for ≥6 months following antibiotic treatment for Lyme disease. Methods Untargeted liquid chromatography–mass spectrometry metabolomics was used to determine longitudinal metabolic responses and biosignatures in PTLDS and clinically cured non-PTLDS Lyme patients. Evaluation of biosignatures included (1) defining altered classes of metabolites, (2) elastic net regularization to define metabolites that most strongly defined PTLDS and non-PTLDS patients at different time points, (3) changes in the longitudinal abundance of metabolites, and (4) linear discriminant analysis to evaluate robustness in a second patient cohort. Results This study determined that observable metabolic differences exist between PTLDS and non-PTLDS patients at multiple time points. The metabolites with differential abundance included those from glycerophospholipid, bile acid, and acylcarnitine metabolism. Distinct longitudinal patterns of metabolite abundance indicated a greater metabolic variability in PTLDS versus non-PTLDS patients. Small numbers of metabolites (6 to 40) could be used to define PTLDS versus non-PTLDS patients at defined time points, and the findings were validated in a second cohort of PTLDS and non-PTLDS patients. Conclusions These data provide evidence that an objective metabolite-based measurement can distinguish patients with PTLDS and help understand the underlying biochemistry of PTLDS.

Immunology↗
DOE OSTI2025

Particle hit clustering and identification using point set transformers in liquid argon time projection chambers

Liquid argon time projection chambers are often used in neutrino physics and dark-matter searches because of their high spatial resolution. The images generated by these detectors are extremely sparse, as the energy values detected by most of the detector are equal to 0, meaning that despite their high resolution, most of the detector is unused in a particular interaction. Instead of representing all of the empty detections, the interaction is usually stored as a sparse matrix, a list of detection locations paired with their energy values. Traditional machine learning methods that have been applied to particle reconstruction such as convolutional neural networks (CNNs), however, cannot operate over data stored in this way and therefore must have the matrix fully instantiated as a dense matrix. Operating on dense matrices requires a lot of memory and computation time, in contrast to directly operating on the sparse matrix. We propose a machine learning model using a point set neural network that operates over a sparse matrix, greatly improving both processing speed and accuracy over methods that instantiate the dense matrix, as well as over other methods that operate over sparse matrices. Compared to competing state-of-the-art methods, our method improves classification performance by 14%, segmentation performance by more than 22%, while taking 80% less time and using 66% less memory. Compared to state-of-the-art CNN methods, our method improves classification performance by more than 86%, segmentation performance by more than 71%, while reducing runtime by 91% and reducing memory usage by 61%.

calibration and fitting methods↗
NASA NTRS2002

Classification image weights and internal noise level estimation

For the linear discrimination of two stimuli in white Gaussian noise in the presence of internal noise, a method is described for estimating linear classification weights from the sum of noise images segregated by stimulus and response. The recommended method for combining the two response images for the same stimulus is to difference the average images. Weights are derived for combining images over stimuli and observers. Methods for estimating the level of internal noise are described with emphasis on the case of repeated presentations of the same noise sample. Simple tests for particular hypotheses about the weights are shown based on observer agreement with a noiseless version of the hypothesis.

Pattern Recognition, Visual/physiology↗
DOE OSTI2024

Efficient distributed continual learning for steering experiments in real-time

Deep learning has emerged as a powerful method for extracting valuable information from large volumes of data. However, when new training data arrives continuously (i.e., is not fully available from the beginning), incremental training suffers from catastrophic forgetting (i.e., new patterns are reinforced at the expense of previously acquired knowledge). Training from scratch each time new training data becomes available would result in extremely long training times and massive data accumulation. Rehearsal-based continual learning has shown promise for addressing the catastrophic forgetting challenge, but research to date has not addressed performance and scalability. To fill this gap, we propose an approach based on a distributed rehearsal buffer that efficiently complements data-parallel training on multiple GPUs to achieve high accuracy, short runtime, and scalability. It leverages a set of buffers (local to each GPU) and uses several asynchronous techniques for updating these local buffers in an embarrassingly parallel fashion, all while handling the communication overheads necessary to augment input minibatches using unbiased, global sampling. We further propose a generalization of rehearsal buffers to support both classification and generative learning tasks, as well as more advanced rehearsal strategies (notably Dark Experience Replay, leveraging knowledge distillation). We illustrate this approach with a real-life HPC streaming application from the domain of ptychographic image reconstruction. Furthermore, we run extensive experiments on up to 128 GPUs of the ThetaGPU supercomputer to compare our approach with baselines representative of training-from-scratch (the upper bound in terms of accuracy) and incremental training (the lower bound). Results show that rehearsal-based continual learning achieves a top-5 validation accuracy close to the upper bound, while simultaneously exhibiting a runtime close to the lower bound.

Asynchronous data management↗
NASA NTRS2021

Extreme Lake-Effect Snow from a GPM Microwave Imager Perspective: Observational Analysis and Precipitation Retrieval Evaluation

This study focuses on the ability of the Global Precipitation Measurement (GPM) passive microwave sensors to detect and provide quantitative precipitation estimates (QPE) for extreme lake-effect snowfall events over the United States lower Great Lakes region. GPM Microwave Imager (GMI) high frequency channels can clearly detect intense shallow convective snowfall events. However, GMI Goddard PROfiling (GPROF) QPE retrievals produce inconsistent results when compared against the Multi-Radar/Multi-Sensor (MRMS) ground-based radar reference dataset. While GPROF retrievals adequately capture intense snowfall rates and spatial patterns of one event, GPROF systematically underestimates intense snowfall rates in another event. Furthermore, GPROF produces abundant light snowfall rates that do not conform with MRMS observations. Ad-hoc precipitation rate thresholds are suggested to partially mitigate GPROF’s overproduction of light snowfall rates. The sensitivity and retrieval efficiency of GPROF to key parameters (2-meter temperature, total precipitable water, and background surface type) used to constrain the GPROF a-priori retrieval database are investigated. Results demonstrate that typical lake-effect snow environmental and surface conditions, especially coastal surfaces, are underpopulated in the database and adversely affect GPROF retrievals. For the two presented case studies, using snow cover a-priori database in the locations of originally deemed as coastline improves retrieval. This study suggests that it is particularly important to have more accurate GPROF surface classifications and better representativeness of the a-priori databases to improve intense lake-effect snow detection and retrieval performance.

Lisa Milani↗
DOE OSTI2024

Data from "A Bayesian Record Linkage Approach to Applications in Tree Demography Using Overlapping LiDAR Scans"

Processed LiDAR data and environmental covariates from 2015 and 2019 LiDAR scans in the Vicinity of Snodgrass Mountain (Western Colorado, USA), in a geographic subset used in primary analysis for the research paper.This package contains LiDAR-derived canopy height maps for 2015 and 2019, crown polygons derived from the height maps using a segmentation algorithm, and environmental covariates supporting the model of forest growth. Source datasets include August 2015 and August 2019 discrete-return LiDAR point clouds collected by Quantum Geospatial for terrain mapping purposes on behalf of the Colorado Hazard Mapping Program and the Colorado Water Conservation Board. Both datasets adhere to the USGS QL2 quality standard. The point cloud data were processed using the R package lidR to generate a canopy height model representing maximum vegetation height above the ground surface, using a pit-free algorithm.This dataset was compiled to assess how spatial patterns of tree growth in montane and subalpine forests are influenced by water and energy availability. Understanding these growth patterns can provide insight into forest dynamics in the Southern Rocky Mountains under changing climatic conditions.This dataset contains .tif, .csv, and .txt files. This dataset additionally includes a file-level metadata (flmd.csv) file that lists each file contained in the dataset with associated metadata; and a data dictionary (dd.csv) file that contains column/row headers used throughout the files along with a definition, units, and data type.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2023

Array-Based Machine Learning for Functional Group Detection in Electron Ionization Mass Spectrometry

Mass spectrometry is a ubiquitous technique capable of complex chemical analysis. The fragmentation patterns that appear in mass spectrometry are an excellent target for artificial intelligence methods to automate and expedite the analysis of data to identify targets such as functional groups. To develop this approach, we trained models on electron ionization (a reproducible hard fragmentation technique) mass spectra so that not only the final model accuracies but also the reasoning behind model assignments could be evaluated. The convolutional neural network (CNN) models were trained on 2D images of the spectra using transfer learning of Inception V3, and the logistic regression models were trained using array-based data and Scikit Learn implementation in Python. Our training dataset consisted of 21,166 mass spectra from the United States’ National Institute of Standards and Technology (NIST) Webbook. The data was used to train models to identify functional groups, both specific (e.g., amines, esters) and generalized classifications (aromatics, oxygen-containing functional groups, and nitrogen-containing functional groups). We found that the highest final accuracies on identifying new data were observed using logistic regression rather than transfer learning on CNN models. It was also determined that the mass range most beneficial for functional group analysis is 0–100 m/z. We also found success in correctly identifying functional groups of example molecules selected from both the NIST database and experimental data. Beyond functional group analysis, we also have developed a methodology to identify impactful fragments for the accurate detection of the models’ targets. The results demonstrate a potential pathway for analyzing and screening substantial amounts of mass spectral data.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

The System for Classification of Low-Pressure Systems (SyCLoPS): An All-In-One Objective Framework for Large-Scale Data Sets

We propose the first unified objective framework (SyCLoPS) for detecting and classifying all types of low-pressure systems (LPSs) in a given data set. We use the state-of-the-art automated feature tracking software TempestExtremes (TE) to detect and track LPS features globally in ERA5 and compute 16 parameters from commonly found atmospheric variables for classification. A Python classifier is implemented to classify all LPSs at once. The framework assigns 16 different labels (classes) to each LPS data point and designates four different types of high-impact LPS tracks, including tracks of tropical cyclone (TC), monsoonal system, subtropical storm and polar low. The classification process involves disentangling high-altitude and drier LPSs, differentiating tropical and non-tropical LPSs using novel criteria, and optimizing for the detection of the four types of high-impact LPS. A comparison of our labels with those in the International Best Track Archive for Climate Stewardship (IBTrACS) revealed an overall accuracy of 95% in distinguishing between tropical systems, extratropical cyclones, and disturbances. SyCLoPS produces a better TC detection skill compared to the previous algorithms, highlighted by an approximately 6% reduction in the false alarm rate compared to the previous TE algorithm. The vertical cross section composite of the four types of high-impact LPS we detect each shows distinct structural characteristics. Finally, we demonstrate that SyCLoPS is valuable for investigating various aspects of LPSs in climate data, such as the evolution of a single LPS track, patterns of LPS frequencies, and precipitation or wind influence associated with a particular LPS class.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Northern Hemisphere Winter Air Temperature Patterns and Their Associated Atmospheric and Ocean Conditions

The Northern Hemisphere (NH) has experienced winter Arctic warming and continental cooling in recent decades, but the dominant patterns in winter surface air temperature (SAT) are not well understood. Here, a self-organizing map (SOM) analysis is performed to identify the leading patterns in winter daily SAT fields from 1979 to 2018, and their associated atmospheric and ocean conditions are also examined. Three distinct winter SAT patterns with two phases of nearly opposite signs and a time scale of 7–12 days are found: one pattern exhibits concurrent SAT anomalies of the same sign over North America (NA) and northern Eurasia, while the other two patterns show SAT anomalies of opposite signs between, respectively, NA and the Bering Sea, and the Kara Sea and East Asia (EA). Winter SAT variations may arise from changes in the SOM frequencies. Specifically, the observed increasing trends of winter cold extremes over NA, central Eurasia, and EA during 1998–2013 can be understood as a result of the increasing occurrences of some specific SAT patterns. These SOMs are closely related to poleward advection of midlatitude warm air and equatorward movements of polar cold airmass. These meridional displacements of cold and warm airmasses cause concurrent anomalies over different regions not only in SAT but also in water vapor and surface downward longwave radiation. Anomalous sea surface temperatures in the tropical Pacific, midlatitude North Pacific, and North Atlantic and anomalous Arctic sea ice concentrations also concur to support and maintain the anomalous atmospheric circulation that causes the SAT anomalies.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Topological Interpretability for Deep Learning

With the growing adoption of AI-based systems across everyday life, the need to understand their decision-making mechanisms is correspondingly increasing. The level at which we can trust the statistical inferences made from AI-based decision systems is an increasing concern, especially in high-risk systems such as criminal justice or medical diagnosis, where incorrect inferences may have tragic consequences. Despite their successes in providing solutions to problems involving real-world data, deep learning (DL) models cannot quantify the certainty of their predictions. These models are frequently quite confident, even when their solutions are incorrect. This work presents a method to infer prominent features in two DL classification models trained on clinical and non-clinical text by employing techniques from topological and geometric data analysis. We create a graph of a model's feature space and cluster the inputs into the graph's vertices by the similarity of features and prediction statistics. We then extract subgraphs demonstrating high-predictive accuracy for a given label. These subgraphs contain a wealth of information about features that the DL model has recognized as relevant to its decisions. We infer these features for a given label using a distance metric between probability measures, and demonstrate the stability of our method compared to the LIME and SHAP interpretability methods. This work establishes that we may gain insights into the decision mechanism of a DL model. This method allows us to ascertain if the model is making its decisions based on information germane to the problem or identifies extraneous patterns within the data.

Spannaus, Adam↗