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 595 records · Page 33
All sourcesNASA NTRSDOE OSTI
NASA NTRS1980

Model error estimation for large space systems

In-flight estimation of large structure model errors may have to be carried out in order to detect inevitable deficiencies in large structure controller/estimator models. These error estimates can most efficiently be obtained by the minimization of a quadratic functional of the model errors and on the subsequent analysis of the resulting optimal model error estimates. An integral operator approach to estimation leads to a geometrical interpretation of the model error estimation process. One of the significant insights gained with this interpretation is that the actual but unknown model errors can be decomposed as the sum of two distinct components that are orthogonal in some sense. One of these components is a so-called minimal error vector that retains many of the significant dynamics of the actual errors. The basic ideas in the model error estimation approach are first set forth with a two-dimensional analogy that has most of the essential features of the general estimation problem. The generalized results are then established and their application to a reference large structure model illustrated.

Rodriguez, G.↗
DOE OSTI2024

Interpretable, extensible linear and symbolic regression models for charge density prediction using a hierarchy of many-body correlation descriptors

Here, density functional theory (DFT) is routinely used to make electronic structure predictions for high-throughput screening of materials and molecules for technologically relevant areas, like the identification of better catalysts, electronic materials, and drug discovery. However, the DFT formalism is limited by (a) its poor (quadratic-to-quartic) scaling, and (b) the need to perform repeated eigenvalue computations of the electronic Hamiltonian as part of its self-consistent field (SCF) iteration procedure to obtain the converged ground state electron density, ρ (r). Approaches that directly predict ρ (r) of a structure with high accuracy can accelerate conventional SCF calculations and can also be used in linearly scaling methods such as orbital-free DFT. To this end, we present a procedure to predict the ground state electron density of molecular and periodic three-dimensional systems directly from the atomic structure with a particular emphasis on physical interpretability. In our framework, ρ (r) is modeled using many-body correlation descriptors that accurately capture the effects of local atomic arrangements in the neighborhood of a grid point. Our use of a linear regression scheme to fit to charge density data enables transparent analysis of the relative contributions of various types of local atomic correlations. By systematically including increasingly complex correlations, our model is shown to accurately predict ρ (r) for a variety of chemically and electronically diverse systems — amorphous Ge, Al(001) slab, crystalline Ga 2 O 3 , molecular benzene, and polyethylene. We then demonstrate a symbolic regression-based protocol to construct easily computable, interpretable features from lower-order correlations that significantly improves our electron density predictions with effectively no increase in the computational cost.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Explainable machine learning for hydrogen diffusion in metals and random binary alloys

Hydrogen diffusion in metals and alloys plays an important role in the discovery of new materials for fuel cell and energy storage technology. While analytic models use hand-selected features that have clear physical ties to hydrogen diffusion, they often lack accuracy when making quantitative predictions. Machine learning models are capable of making accurate predictions, but their inner workings are obscured, rendering it unclear which physical features are truly important. To develop interpretable machine learning models to predict the activation energies of hydrogen diffusion in metals and random binary alloys, we create a database for physical and chemical properties of the species and use it to fit six machine learning models. Our models achieve root-mean-squared errors between 98–119 meV on the testing data and accurately predict that elemental Ru has a large activation energy, while elemental Cr and Fe have small activation energies. By analyzing the feature importances of these fitted models, we identify relevant physical properties for predicting hydrogen diffusivity. While metrics for measuring the individual feature importances for machine learning models exist, correlations between the features lead to disagreement between models and limit the conclusions that can be drawn. Instead grouped feature importance, formed by combining the features via their correlations, agree across the six models and reveal that the two groups containing the packing factor and electronic specific heat are particularly significant for predicting hydrogen diffusion in metals and random binary alloys. In conclusion, this framework allows us to interpret machine learning models and enables rapid screening of new materials with the desired rates of hydrogen diffusion.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Linear Time-Invariant Models of a Large Cumulus Ensemble

Abstract Methods in system identification are used to obtain linear time-invariant state-space models that describe how horizontal averages of temperature and humidity of a large cumulus ensemble evolve with time under small forcing. The cumulus ensemble studied here is simulated with cloud-system-resolving models in radiative–convective equilibrium. The identified models extend steady-state linear response functions used in past studies and provide accurate descriptions of the transfer function, the noise model, and the behavior of cumulus convection when coupled with two-dimensional gravity waves. A novel procedure is developed to convert the state-space models into an interpretable form, which is used to elucidate and quantify memory in cumulus convection. The linear problem studied here serves as a useful reference point for more general efforts to obtain data-driven and interpretable parameterizations of cumulus convection.

Meteorology & Atmospheric Sciences↗
DOE OSTI2018

Data-Driver Interpretation of Ultradeep Azimuthal Propagation

We report the interpretation of geological features around the borehole stresses on the search of true earth model parameters. Different from the conventional logging-while-drilling (LWD) technologies, such as wireline induction and propagation resistivity tools, the azimuthal resistivity measurement acquires the ability of directional sensitivity. Within this decade, this type of tools has been greatly enhanced with over hundred feet depth-of-detection (DoD). To resolve the interpretation of geological features within 100 feet from the borehole, the multiple-layer structure must be considered. However, the traditional methods can never handle the highly nonlinear inverse problems especially facing on the complicated earth structure. Meanwhile, no justifiable uncertainty of inverse parameters can be provided. The statistical methods have been adopted to search for the global statistical distribution and uncertainties of earth model parameters. However, the most attempts are still stuck in non-uniqueness of the solutions since the structure of earth model has been predefined, where the number of layers is assumed, which introduces another ambiguity into the model interpretation. This paper develops a general trans-dimensional Bayesian methodology to solve non-linear inverse problem of earth layer resistivity and depth to the layer boundaries, from the interpretation of ultra-deep azimuthal resistivity measurements. The presented work is a data-driven approach to search out the multi-dimensional parameter spaces. Finally, the simulations tell that the algorithm can efficiently solve the earth model interpretation within a hundred feet from borehole, and examine the uncertainty of earth model parameters.

58 GEOSCIENCES↗
NASA NTRS1977

Practical quantitative fringe interpretation in a composite mobile holographic nondestructive test system

A simplified theoretical model for the interpretation of the double-exposure holographic interference fringe loci due to the general three-dimensional displacements has been derived for the specific composite mobile holographic nondestructive test system. The model, representing a good approximation to a more tedious theoretical result, predicts that a combination of in-plane and out-of-plane displacements of the surface will produce concentric circular-shaped fringe patterns.

Liu, H.-K.↗
DOE OSTI2021

PayamDiba/CoNSEPT

Deciphering the sequence-function relationship encoded in enhancers holds the key to interpreting non-coding variants and understanding mechanisms of transcriptomic variation. Several quantitative models exist for predicting enhancer function and underlying mechanisms; however, there has been no systematic comparison of these models characterizing their relative strengths and shortcomings. Here, we interrogated a rich data set of neuroectodermal enhancers in Drosophila, representing cis- and trans- sources of expression variation, with a suite of biophysical and machine learning models. We performed rigorous comparisons of thermodynamics-based models implementing different mechanisms of activation, repression and cooperativity. Moreover, we developed a convolutional neural network (CNN) model, called CoNSEPT, that learns enhancer ‘grammar’ in an unbiased manner. CoNSEPT is the first general-purpose CNN tool for predicting enhancer function in varying conditions, such as different cell types and experimental conditions, and we show that such complex models can suggest interpretable mechanisms. We found model-based evidence for mechanisms previously established for the studied system, including cooperative activation and short-range repression. The data also favored one hypothesized activation mechanism over another and suggested an intriguing role for a direct, distance-independent repression mechanism. Our modeling shows that while fundamentally different models can yield similar fits to data, they vary in their utility for mechanistic inference.

Dibaeinia, Payam↗
DOE OSTI2020

Multidimensional simulations of Mckenna-driven flow tube configuration: Investigating non-ideality in NO x formation flow tube experiments

Multidimensional simulations have been conducted to simulate atmospheric pressure, flat-flame/McKenna-burner-driven-flow tube experiments targeted to obtain NO x speciation data for predicting/analyzing syngas combustion emissions. In a prior work, we demonstrated the impacts of multidimensional transport on post flame region prediction departures from those assuming unidimensional flow/transport conditions. In this work, we develop and utilize a multidimensional laminar reacting flow solver to simulate the fully coupled flame and post flame regions to further elucidate the impacts of the earlier unidimensional modeling assumptions on interpreting post flame NO x experimental data. The model is used to simulate a lean, premixed syngas/air flame and its associated post flame regions within a cylindrical flow-tube-like arrangement. The combustion process takes place under atmospheric condition with trace amount of NO x seeding fed into the inlet gas stream. The spatial evolution of NO x species (NO and NO 2 ) in the flame and in the post-combustion zone suggests two distinct regions: 1) a region encompassing the flame structure itself; and 2) a post flame region in which the temperature decays due to both axial and radial transport processes. The predictions show that for the conditions studied, a pulsatile flow field exists due to the formation of an expanding and contracting recirculation zone in the outer periphery of the flow tube. By resolving the nature of the flow, the resulting time-averaged temperature and species concentrations show improved agreement with existing experimental measurements. The flow-field interaction results in radial inhomogeneities in the NO 2 profiles with the maximum concentration offset from the flow centerline. The location of the peak in NO 2 is coupled with radial temperature gradients from wall cooling effects and their significant influence on NO/NO 2 interconversion kinetics, producing notable NO 2 accumulation in regions near the wall. Geometrical configurations capable of suppressing/minimizing the pulsatile nature are also investigated and the results are compared. Other experimental configurations could be considered in parametric simulations to determine the optimal configuration that would minimize non-idealities in the observations. The work shows the value in performing such computations in advance of settling on a particular design for flow tube/flow reactor experiments.

42 ENGINEERING↗
DOE OSTI2022

Physics-Informed AI for Climate and Weather Risk Prediction (Final Report)

As part of the work Terrafuse developed 1) a model for wildfire risk in California, and 2) a model for downscaled wind fields from Numerical Weather Prediction (NWP) numerical models. The wildfire model is based on 20 years of historical data and captures the dependence of wildfire incidence and spread on climatic, weather and land-use variables by training a machine learning model. Nonlinear relationships between input features are learned and expressed by the model and model transparency allows features to be ranked and interpreted. The fire model is of use for accurately predicting real-time daily and long-term wildfire risk for use cases in energy and insurance. The downscaled wind model is a spatio-temporal deep learning model that emulates the influence of high-resolution variables on wind speed, allowing coarse-resolution operational NWP models to be accurately expressed on fine grids at high resolution, with application to wind energy and weather forecasting.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Development of an Integrated Multidiagnostic to Assess the High-Z Impurity Fluxes in the Metallic Environment of WEST Using IMAS

WEST is an actively cooled, long-pulse tokamak with nearly all plasma-facing components (PFC) made of tungsten. One of the aims of WEST is to study plasma operations with tungsten PFCs in preparation for long-pulse operations on high-Z divertor devices, such as ITER. For long-pulse operation, the high-Z impurity content and transport to the core plasma are critical concerns that require further measurement and interpretation to improve plasma performance and PFC durability. This work details the impurity influxes in WEST during a series of discharges in which the lower hybrid (LH) injected power was incrementally increased. An analysis has been performed of measurements collected from an array of edge diagnostics. Visible spectroscopy was utilized to measure the spectral radiances generated by fuel particles (D) and impurities (W, O, and C) at the divertor and at the antennas with a newly developed spectral peak-fitting tool used to analyze the data in WEST. The scrape-off layer (SOL) plasma conditions (n e and T e ) measured at the divertor target with flush Langmuir probes and near the outer mid-plane (OMP) using reciprocating Langmuir probes (RCPs) are used to evaluate the number of ionizations per photons (S/XB) coefficients required to estimate the impurity fluxes obtained with the collisional-radiative model ColRadPy. Here, the array of edge diagnostics discussed in this work, coupled with SOL plasma modeling tools, represents a multidiagnostic interpretative modeling workflow that will continue to be applied to upcoming experimental campaigns on the WEST experiment to assess high-Z impurity transport.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Multimodel ensemble predictions of precipitation using bayesian neural networks

Multimodel ensembling improves predictions and considers model uncertainties. In this study, we present a Bayesian Neural Network (BNN) ensemble approach for large-scale precipitation predictions based on a set of climate models. BNN infers spatiotemporally varying model weights and biases through the calibration against observations. This ensemble scheme of BNN sufficiently leverages individual model skill for accurate predictions as well as provides interpretability about which models contribute more to the ensemble prediction at which locations and times to inform model development. Additionally, BNN accurately quantifies epistemic uncertainty to avoid overconfident projections. We demonstrate BNN’s superior prediction performance to three state-of-the-art ensemble approaches and discuss its interpretability and uncertainty quantification.

Fan, Patrick↗
DOE OSTI2022

Multimodel Ensemble Predictions of Precipitation using Bayesian Neural Networks

Multimodel ensembling improves predictions and considers model uncertainties. In this study, we present a Bayesian Neural Network (BNN) ensemble approach for large-scale precipitation predictions based on a set of climate models. BNN infers spatiotemporally varying model weights and biases through the calibration against observations. This ensemble scheme of BNN sufficiently leverages individual model skill for accurate predictions as well as provides interpretability about which models contribute more to the ensemble prediction at which locations and times to inform model development. Additionally, BNN accurately quantifies epistemic uncertainty to avoid overconfident projections. We demonstrate BNN’s superior prediction performance to three state-of-the-art ensemble approaches and discuss its interpretability and uncertainty quantification.

Fan, Patrick↗
DOE OSTI2022

Search for electroweak production of charginos and neutralinos in proton-proton collisions at $$ \sqrt{s} $$ = 13 TeV

A direct search for electroweak production of charginos and neutralinos is presented. Events with three or four leptons, with up to two hadronically decaying $\tau$ leptons, or two same-sign light leptons are analyzed. The data sample consists of 137 fb$^{-1}$ of proton-proton collisions with a center of mass energy of 13 TeV, recorded with the CMS detector at the LHC. The results are interpreted in terms of several simplified models. These represent a broad range of production and decay scenarios for charginos and neutralinos. A parametric neural network is used to target several of the models with large backgrounds. In addition, results using orthogonal search regions are provided for all the models, simplifying alternative theoretical interpretations of the results. Depending on the model hypotheses, charginos and neutralinos with masses up to values between 300 and 1450 GeV are excluded at 95% confidence level.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Studies of new Higgs boson interactions through nonresonant $HH$ production in the $ b\overline{b}\gamma \gamma $ final state in $pp$ collisions at $ \sqrt{s} $ = 13 TeV with the ATLAS detector

A search for nonresonant Higgs boson pair production in the $ b\overline{b}\gamma \gamma $ final state is performed using 140 fb –1 of proton-proton collisions at a centre-of-mass energy of 13 TeV recorded by the ATLAS detector at the CERN Large Hadron Collider. This analysis supersedes and expands upon the previous nonresonant ATLAS results in this final state based on the same data sample. The analysis strategy is optimised to probe anomalous values not only of the Higgs (H) boson self-coupling modifier κλ but also of the quartic HHVV (V = W, Z) coupling modifier κ 2V . No significant excess above the expected background from Standard Model processes is observed. An observed upper limit μHH < 4.0 is set at 95% confidence level on the Higgs boson pair production cross-section normalised to its Standard Model prediction. The 95% confidence intervals for the coupling modifiers are –1.4 < κ λ < 6.9 and –0.5 < κ 2V < 2.7, assuming all other Higgs boson couplings except the one under study are fixed to the Standard Model predictions. The results are interpreted in the Standard Model effective field theory and Higgs effective field theory frameworks in terms of constraints on the couplings of anomalous Higgs boson (self-)interactions.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Search for a scalar partner of the top quark in the all-hadronic $t{\bar{t}}$ plus missing transverse momentum final state at $\sqrt{s}=13$ TeV with the ATLAS detector

A search for direct pair production of scalar partners of the top quark (top squarks or scalar third-generation up-type leptoquarks) in the all-hadronic $t{\bar{t}}$ plus missing transverse momentum final state is presented. The analysis of 139 $\hbox {fb}^{-1}$ of ${\sqrt{s}=13}$ TeV proton–proton collision data collected using the ATLAS detector at the LHC yields no significant excess over the Standard Model background expectation. To interpret the results, a supersymmetric model is used where the top squark decays via ${\tilde{t}} \rightarrow t^{(*)} {\tilde{\chi }}^0_1$ t ~ → t ( * ) χ ~ 1 0 , with $t^{(*)}$ t ( * ) denoting an on-shell (off-shell) top quark and ${\tilde{\chi }}^0_1$ χ ~ 1 0 the lightest neutralino. Three specific event selections are optimised for the following scenarios. In the scenario where $m_{{\tilde{t}}}> m_t+m_{{\tilde{\chi }}^0_1}$ m t ~ > m t + m χ ~ 1 0 , top squark masses are excluded in the range 400–1250 GeV for ${\tilde{\chi }}^0_1$ χ ~ 1 0 masses below 200 GeV at 95% confidence level. In the situation where $m_{{\tilde{t}}}\sim m_t+m_{{\tilde{\chi }}^0_1}$ m t ~ ~ m t + m χ ~ 1 0 , top squark masses in the range 300–630 GeV are excluded, while in the case where $m_{{\tilde{t}}}< m_W+m_b+m_{{\tilde{\chi }}^0_1}$ m t ~ < m W + m b + m χ ~ 1 0 (with $m_{{\tilde{t}}}-m_{{\tilde{\chi }}^0_1}\ge 5$ m t ~ - m χ ~ 1 0 ≥ 5 GeV), considered for the first time in an ATLAS all-hadronic search, top squark masses in the range 300–660 GeV are excluded. Limits are also set for scalar third-generation up-type leptoquarks, excluding leptoquarks with masses below 1240 GeV when considering only leptoquark decays into a top quark and a neutrino.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS1975

Local time variations of particle flux produced by an electrostatic field in the magnetosphere

The local time variations of particle distributions, which can be produced by the sudden enhancement of a cross-tail electric field, are examined by using adiabatic theory. The strong coupling between acceleration, inward motion, and east-west drift is emphasized. The model is used to interpret recent measurements of delay times of newly accelerated electrons at different local times at geostationary orbit. Observations of the asymmetric storm time 'ring current' are interpreted in terms of the model without invoking particle loss.

Kivelson, M. G.↗
NASA NTRS1997

Modeling and Data Analysis of SONEX Observations

Primary focus of the study is on the sources, photochemistry and transport of reactive nitrogen species, O3, as well as other O3 precursors such as NMHC and CO. Major objectives are: (1) Identify the major sources of NO(x) and quantify the contribution of each source to the NO(x) distribution in the free troposphere, particularly in the upper troposphere. (2) Understand the odd hydrogen photochemistry and the recycling of NO(x) in the free troposphere. (3) Evaluate the photochemical production and destruction of O3. A combination of modeling and data analysis/interpretation approach is used. We use a three-dimensional model and a box model to simulate, analyze, and interpret the observations. Both models are needed because they are complementary and each has its advantages and limitations. Proven analytical techniques, including correlation of NOx with various tracers and with different air masses, are applied to both observations and model results to deduce useful information toward accomplishing the objectives. In addition, the SONEX observations are compared and analyzed in a systematic way against other airborne experiments conducted over remote atmosphere, especially those from NASA GTE experiments, AASE 1 and 2, and SUCCESS.

Liu, Shaw C.↗