Search NASA⌕ Search

SEARCH · Search NASA

Results for “Model interpretation”

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 199 records · Page 11

A primer on artificial intelligence in plant digital phenomics: embarking on the data to insights journey

Artificial intelligence (AI) has emerged as a fundamental component of global agricultural research that is poised to impact on many aspects of plant science. In digital phenomics, AI is capable of learning intricate structure and patterns in large datasets. We provide a perspective and primer on AI applications to phenome research. We propose a novel human-centric explainable AI (X-AI) system architecture consisting of data architecture, technology infrastructure, and AI architecture design. We clarify the difference between post hoc models and 'interpretable by design' models. We include guidance for effectively using an interpretable by design model in phenomic analysis. We also provide directions to sources of tools and resources for making data analytics increasingly accessible. In conclusion, this primer is accompanied by an interactive online tutorial.

60 APPLIED LIFE SCIENCES↗

Explaining drivers of housing prices with nonlinear hedonic regressions

Housing markets play a critical role in shaping the spatial and demographic evolution of urban areas. Simulating housing price dynamics can enhance projections of future urban development outcomes. However, traditional hedonic regressions for housing prices, which neglect nonlinear interactions among explanatory variables, often exhibit limited predictive performance. While machine learning (ML) methods can provide a more flexible representation of the relationships between predictors, they are often regarded as “black boxes” due to their complexity and lack of transparency. Interpretable ML techniques provide a promising route by combining the flexibility of ML methods with approaches to analyze the relationships between inputs and outputs. In this study, we employ interpretable ML to analyze the patterns driving the housing market in Baltimore, Maryland, USA. We train an Artificial Neural Network (ANN) to predict Baltimore housing prices based on structural characteristics (e.g., home size, number of stories) and locational attributes (e.g., distance to the city center). We then conduct sensitivity and Partial Dependence Plot (PDP) analyses to interpret the fitted ANN model. We find that the ML model achieves higher predictive accuracy and explains 16 % more of housing price variance than a traditional linear regression model. The interpretable ML model also reveals more nuanced and realistic nonlinear relationships between housing sales price and predictors as well as interactive effects underlying Baltimore home price dynamics. For instance, while the linear model indicates a steady housing price increase over time, our interpretable ML model detects a post-2008 decline, with smaller properties experiencing the sharpest drop.

97 MATHEMATICS AND COMPUTING↗

Learning to Branch with Interpretable Machine Learning Models

Machine learning is being increasingly used in improving decisions made within branch-and-bound algorithms for solving mixed-integer programs (MIPs). Branching is a key component in branch-and-bound algorithms, this work presents IDAES-core project update on building simple and interpretable machine learning models for branching and improving decision-making tools applied for the optimization of advanced energy systems.

Bayramoglu, Selin↗

Interpretable Machine Learning Models for Practical Antimonate Electrocatalyst Performance

Computationally predicting the performance of catalysts under reaction conditions is a challenging task due to the complexity of catalytic surfaces and their evolution in situ, different reaction paths, and the presence of solid-liquid interfaces in the case of electrochemistry. We demonstrate here how relatively simple machine learning models can be found that enable prediction of experimentally observed onset potentials. Inputs to our model are comprised of data from the oxygen reduction reaction on non-precious transition-metal antimony oxide nanoparticulate catalysts with a combination of experimental conditions and computationally affordable bulk atomic and electronic structural descriptors from density functional theory simulations. From human-interpretable genetic programming models, we identify key experimental descriptors and key supplemental bulk electronic and atomic structural descriptors that govern trends in onset potentials for these oxides and deduce how these descriptors should be tuned to increase onset potentials. Here, we finally validate these machine learning predictions by experimentally confirming that scandium as a dopant in nickel antimony oxide leads to a desired onset potential increase. Macroscopic experimental factors are found to be crucially important descriptors to be considered for models of catalytic performance, highlighting the important role machine learning can play here even in the presence of small datasets.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Interpretable machine learning models classify minerals via spectroscopy

Developing methods to identify mineral species confidently and rapidly from Raman spectral analysis is critical to numerous fields. Traditionally, analysis relies on pattern matching the Raman spectrum of an unknown dataset with a supporting library of well-characterized spectral data, which may prove difficult for environmental samples that are poorly crystalline or phase mixtures. Here, we developed interpretable machine learning models that can classify uranium minerals by secondary oxyanion chemistry and other physicochemical properties based solely on Raman spectra. This new ML method produces a mineral profile of physical and chemical properties for an unknown sample and can rapidly classify or identify unknown minerals from Raman data, without the need for an exact pattern match in a spectral library. Training models are validated by 1. Strong correlation of high confidence model regions with published spectroscopic assignments and 2. Correct classification of a mineral not present in training data. Training data are from the Compendium of Uranium Raman and Infrared Experimental Spectra and available crystallographic information files within the open-source Smart Spectral Matching scientific framework. Physically meaningful classifier models can rapidly identify key structural and chemical information about unknown uranium minerals and the overall methodology is broadly applicable for mineral phases.

Machine learning↗

Modified equilibrium-dispersive model for the interpretation of the efficiency of columns packed on core-shell particle

A modified Equilibrium Dispersive (ED) Model is proposed for the modeling of chromatographic processes in columns packed with shell-particle adsorbents and operated under very high pressures. This new model was validated on the basis of experimental results obtained with 2.1 mm x 150 mm columns packed with superficially porous 1.7 {micro}m Kinetex-C{sub 18} particles and with classical columns packed with 1.7 {micro}m BEH-C{sub 18} fully porous particles. The influence of the heat friction on the performance of these columns was analyzed by comparing the experimental and calculated peak profiles. Moreover a theoretical analysis of the influence the solid-core conductivity on the column efficiency was discussed.

36 MATERIALS SCIENCE↗

Methodology for interpretable reinforcement learning model for HVAC energy control

Deep reinforcement learning (DRL) approaches have been used in various application areas to improve efficiency, optimization, or automation. However, very little is known about how the DRL algorithms make decisions and what features affect their performance. Using a case study of a DRL based Heating, Ventilation and Air Conditioning (HVAC) optimization methodology, we demonstrate how we can address these challenges by applying interpretability tools and systematically exploring the model inputs for better understanding the DRL behaviour and decision making process. We developed a methodology for interpretable reinforcement learning and evaluated our approach in real-world house located in Knoxville, TN. Our findings explain the reasoning behind DRL-based optimization decisions under different circumstances which has been discussed and confirmed by the experts in the field.

Kotevska, Olivera↗

From deep TLS validation to ensembles of atomic models built from elemental motions. II. Analysis of TLS refinement results by explicit interpretation

TLS modelling was developed by Schomaker and Trueblood to describe atomic displacement parameters through concerted (rigid-body) harmonic motions of an atomic group [Schomaker & Trueblood (1968), Acta Cryst. B24, 63–76]. The results of a TLS refinement are T, L and S matrices that provide individual anisotropic atomic displacement parameters (ADPs) for all atoms belonging to the group. These ADPs can be calculated analytically using a formula that relates the elements of the TLS matrices to atomic parameters. Alternatively, ADPs can be obtained numerically from the parameters of concerted atomic motions corresponding to the TLS matrices. Both procedures are expected to produce the same ADP values and therefore can be used to assess the results of TLS refinement. Here, the implementation of this approach in PHENIX is described and several illustrations, including the use of all models from the PDB that have been subjected to TLS refinement, are provided.

59 BASIC BIOLOGICAL SCIENCES↗

TRANSP integrated modeling code for interpretive and predictive analysis of tokamak plasmas

This paper provides a comprehensive review of the TRANSP code, a sophisticated tool for interpretive and predictive analysis of tokamak plasmas, detailing its major capabilities and features. It describes the equations for particle, power, and momentum balance analysis, as well as the poloidal field diffusion equations. The paper outlines the spatial and time grids used in TRANSP and details the equilibrium assumptions and solvers. Various models for heating and current drive and radiation, including updates to the NUBEAM model, are discussed. The handling of large-scale events such as sawtooth crashes and pellet injections is examined, along with the predictive capabilities for advancing plasma profiles. The integration of TRANSP with the ITER Integrated Modeling and Analysis Suite (IMAS) is highlighted, demonstrating enhanced data access and analysis capabilities. Additionally, the paper discusses best practices and continuous integration techniques to enhance TRANSP's robustness. The suite of TRANSP tools, designed for efficient data analysis and simulation, further supports the optimization of tokamak operations and coupling with other tokamak codes. Continuous development and support ensure that TRANSP remains a major code for the analysis of experimental data for controlled thermonuclear fusion, maintaining its critical role in supporting the optimization of tokamak operations and advancing fusion research.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An accretion-jet model for M87: interpreting the spectral energy distribution and faraday rotation measure

M87 is arguably the best supermassive black hole (BH) to explore jet and/or accretion physics, due to its proximity and fruitful high-resolution multi-waveband observations. We model the multi-wavelength spectral energy distribution (SED) of the M87 core that observed at a scale of 0.4 arcsec (∼10{sup 5} R {sub g}, R {sub g} is gravitational radius), as recently presented by Prieto et al. Similar to Sgr A*, we find that the millimeter bump as observed by the Atacama Large Millimeter/submillimeter Array can be modeled by the synchrotron emission of the thermal electrons in an advection-dominated accretion flow (ADAF), while the low-frequency radio emission and X-ray emission may predominantly come from the jet. The millimeter radiation from ADAF predominantly comes from the region within 10R {sub g}, which is roughly consistent with the recent very long baseline interferometry observations at 230 GHz. We further calculate the Faraday rotation measure (RM) from both ADAF and jet models, and find that the RM predicted from the ADAF is roughly consistent with the measured value, while the RM predicted from the jet is much higher if jet velocity close to the BH is low or moderate (e.g., v {sub jet} ≲ 0.6 c). With the constraints from the SED modeling and RM, we find that the accretion rate close to the BH horizon is ∼(0.2--1) ×10{sup −3} M{sub ⊙} yr{sup −1}≪ M-dot {sub B}∼0.2 M{sub ⊙} yr{sup −1} ( M-dot {sub B} is Bondi accretion rate), where the electron density profile, n {sub e} ∝ r {sup ∼−1}, in the accretion flow, is consistent with that determined from X-ray observation inside the Bondi radius and recent numerical simulations.

79 ASTRONOMY AND ASTROPHYSICS↗

A classical mechanics model for the interpretation of piezoelectric property data

In order to provide a means of understanding, the relationship between the primary electromechanical coefficients and simple crystal chemistry parameters for piezoelectric materials, a static analysis of a 3 atom, dipolar molecule has been undertaken to derive relationships for elastic compliance s{sup E}, dielectric permittivity ε{sup X}, and piezoelectric charge coefficient d in terms of an effective ionic charge and two inter-atomic force constants. The relationships demonstrate the mutual interdependence of the three coefficients, in keeping with experimental evidence from a large dataset of commercial piezoelectric materials. It is shown that the electromechanical coupling coefficient k is purely an expression of the asymmetry in the two force constants or bond compliances. The treatment is extended to show that the quadratic electrostriction relation between strain and polarization, in both centrosymmetric and non-centrosymmetric systems, is due to the presence of a non-zero 2nd order term in the bond compliance. Comparison with experimental data explains the counter-intuitive, positive correlation of k with s{sup E} and ε{sup X} and supports the proposition that high piezoelectric activity in single crystals is dominated by large compliance coupled with asymmetry in the sub-cell force constants. However, the analysis also shows that in polycrystalline materials, the dielectric anisotropy of the constituent crystals can be more important for attaining large charge coefficients. The model provides a completely new methodology for the interpretation of piezoelectric and electrostrictive property data and suggests methods for rapid screening for high activity in candidate piezoelectric materials, both experimentally and by novel interrogation of ab initio calculations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Increased Interpretability for Model-Driven Deception: MARS LDRD Project

Machine learning has been proposed as a solution to several cybersecurity solutions and one of the most promising applications is for digital twins for intrusion detection and driving deceptive defense. However, machine learning techniques often result in a black-box function that is difficult for end users to interpret which for deception limits their ability to effectively define decoys. In this report, an approach to validate the equations learned are accurate is provided and demonstrated. Following, begins the process of addressing this issue for a model-driven deception technology that produces equations representing the physical process controlled by operation technology devices. This research was performed by applying subject matter expert context to machine learned models.

97 MATHEMATICS AND COMPUTING↗

Development of a New Chelation Model: Bioassay Data Interpretation and Dose Assessment after Plutonium Intake via Wound and Treatment with DTPA

The administration of chelation therapy to treat significant intakes of actinides, such as plutonium, affects the actinide’s normal biokinetics. In particular, it enhances the actinide’s rate of excretion, such that the standard biokinetic models cannot be applied directly to the chelation-affected bioassay data in order to estimate the intake and assess the radiation dose. Here we propose a new chelation model that can be applied to the chelation-affected bioassay data after plutonium intake via wound and treatment with DTPA. In the proposed model, chelation is assumed to occur in the blood, liver, and parts of the skeleton. Ten datasets, consisting of measurements of 14 C-DTPA, 238 Pu, and 239 Pu involving humans given radiolabeled DTPA and humans occupationally exposed to plutonium via wound and treated with chelation therapy, were used for model development. The combined dataset consisted of daily and cumulative excretion (urine and feces), wound counts, measurements of excised tissue, blood, and post-mortem tissue analyses of liver and skeleton. The combined data were simultaneously fit using the chelation model linked with a plutonium systemic model, which was linked to an ad hoc wound model. The proposed chelation model was used for dose assessment of the wound cases used in this study.

60 APPLIED LIFE SCIENCES↗

THE ORBITAL PHASES AND SECONDARY TRANSITS OF KEPLER-10b. A PHYSICAL INTERPRETATION BASED ON THE LAVA-OCEAN PLANET MODEL

The Kepler mission has made an important observation: the first detection of photons from a terrestrial planet by observing its phase curve (Kepler-10b). This opens a new field in exoplanet science: the possibility of obtaining information about the atmosphere and surface of rocky planets, objects of prime interest. In this Letter, we apply the Lava-ocean model to interpret the observed phase curve. The model, a planet without atmosphere and a surface partially made of molten rocks, has been proposed for planets of the class of CoRoT-7b, i.e., rocky planets very close to their star (at a few stellar radii). Kepler-10b is a typical member of this family. It predicts that the light from the planet has an important emission component in addition to the reflected one, even in the Kepler spectral band. Assuming an isotropical reflection of light by the planetary surface (Lambertian-like approximation), we find that a Bond albedo of {approx}50% can account for the observed amplitude of the phase curve, as opposed to a first attempt where an unusually high value was found. We propose a physical process to explain this still large value of the albedo. The overall interpretation can be tested in the future with instruments such as the James Webb Space Telescope or the Exoplanet Characterization Observatory. Our model predicts a spectral dependence that is clearly distinguishable from that of purely reflected light and from that of a planet at a uniform temperature.

79 ASTRONOMY AND ASTROPHYSICS↗

Poster Abstract: Leveraging Large Language Models to Reveal Interpretable Cooling Behaviors from Smart Thermostat Data

Frequent heatwaves and hot summers increasingly challenge occupant comfort, health, and energy grid stability. Addressing these challenges requires a detailed understanding of household cooling behaviors, such as thermostat adjustments and adaptive responses to extreme conditions. Traditional analyses often rely on aggregated numerical metrics that overlook subtle but important household-specific variations. In this study, we introduce a generalizable methodology that integrates large language models (LLMs) with vision capabilities to enable scalable and detailed analysis of residential thermostat data. Using Ecobee's Donate Your Data (DYD) dataset—which provides five-minute records of indoor temperatures, thermostat setpoints, and HVAC runtimes—we focus on two U.S. cities with contrasting summer climates : Austin (TX) and Phoenix (AZ). Because raw time-series data are not well suited for direct LLM analysis, we transform them into visual representations, such as daily indoor temperature trajectories and weekly runtime histograms, to better capture behavioral variations. Leveraging LLMs' visual interpretation, we extract descriptive behavioral features, including temperature preferences, time-of-day cooling orientation, anticipatory versus reactive heatwave responses, and behavioral consistency. These semantic features support unsupervised clustering to identify distinct occupant archetypes at scale, revealing differences—such as morning-centric anticipatory coolers versus households that shift toward warmer setpoints during heatwaves—that can inform demand response, resilience planning, and health-aware interventions. By converting raw numerical data into interpretable behavioral patterns, this methodology enables scalable and practical analysis of occupant behavior, supporting actionable insights for comfort, resilience, and energy management.

Nihar, Kopal↗

DDxNet: a deep learning model for automatic interpretation of electronic health records, electrocardiograms and electroencephalograms

Effective patient care mandates rapid, yet accurate, diagnosis. With the abundance of non-invasive diagnostic measurements and electronic health records (EHR), manual interpretation for differential diagnosis has become time-consuming and challenging. This has led to wide-spread adoption of AI-powered tools, in pursuit of improving accuracy and efficiency of this process. While the unique challenges presented by each modality and clinical task demand customized tools, the cumbersome process of making problem-specific choices has triggered the critical need for a generic solution to enable rapid development of models in practice. In this spirit, we develop DDxNet, a deep architecture for time-varying clinical data, which we demonstrate to be well-suited for diagnostic tasks involving different modalities (ECG/EEG/EHR), required level of characterization (abnormality detection/phenotyping) and data fidelity (single-lead ECG/22-channel EEG). Using multiple benchmark problems, we show that DDxNet produces high-fidelity predictive models, and sometimes even provides significant performance gains over problem-specific solutions.

60 APPLIED LIFE SCIENCES↗