Search NASA⌕ Search

SEARCH · Search NASA

Results for “Model Inference”

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 649 records · Page 36
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Impact of cross section uncertainties on NOvA oscillation analyses

The NOvA experiment is an off-axis long-baseline neutrino oscillation experiment measuring $\nu_{\mu}$ disappearance and $\nu_{e}$ appearance in a $\nu_{\mu}$ beam originating at Fermilab. Oscillations are observed in the far detector relative to measurements made in the near detector, with a full simulation of the beam line and detectors being used to perform extrapolation between them. The neutrino simulation uses the GENIE event generator, which contains implementations of numerous theoretical neutrino interaction models applied to a variety of nuclear targets. However, recent data, recent reanalysis of extant data, and continued development of theoretical models have brought to light deficiencies in the default GENIE cross section model, which informs the predicted spectra used to infer oscillation parameters. We explore how uncertainties in this model, together with modifications to GENIE version 2.12.2 based on external information and NOvA Near Detector (ND) data, affect NOvA's oscillation parameter inferences. We also discuss how these uncertainties can be mitigated by judicious analysis design, including usage of the ND data.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

FedADMP: A Joint Anomaly Detection and Mobility Prediction Framework via Federated Learning

With the proliferation of mobile devices and smart cameras, detecting anomalies and predicting their mobility are critical for enhancing safety in ubiquitous computing systems. Due to data privacy regulations and limited communication bandwidth, it is infeasible to collect, transmit, and store all data from mobile devices at a central location. To overcome this challenge, we propose FedADMP, a federated learning based joint Anomaly Detection and Mobility Prediction framework. FedADMP adaptively splits the training process between the server and clients to reduce computation loads on clients. To protect the privacy of user data, clients in FedADMP upload only intermediate model parameters to the cloud server. We also develop a differential privacy method to prevent the cloud server and external attackers from inferring private information during the model upload procedure. Extensive experiments using real-world datasets show that FedADMP consistently outperforms existing methods.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

The contributions from the progenitor genomes of the mesopolyploid Brassiceae are evolutionarily distinct but functionally compatible

The members of the tribe Brassiceae share a whole-genome triplication (WGT), and one proposed model for its formation is a two-step pair of hybridizations producing hexaploid descendants. However, evidence for this model is incomplete, and the evolutionary and functional constraints that drove evolution after the hexaploidy are even less understood. Here, we report a new genome sequence of Crambe hispanica , a species sister to most sequenced Brassiceae. Using this new genome and three others that share the hexaploidy, we traced the history of gene loss after the WGT using the Polyploidy Orthology Inference Tool (POInT). We confirm the two-step formation model and infer that there was a significant temporal gap between those two allopolyploidizations, with about a third of the gene losses from the first two subgenomes occurring before the arrival of the third. We also, for the 90,000 individual genes in our study, make parental subgenome assignments, inferring, with measured uncertainty, from which of the progenitor genomes of the allohexaploidy each gene derives. We further show that each subgenome has a statistically distinguishable rate of homoeolog losses. There is little indication of functional distinction between the three subgenomes: the individual subgenomes show no patterns of functional enrichment, no excess of shared protein-protein or metabolic interactions between their members, and no biases in their likelihood of having experienced a recent selective sweep. We propose a "mix and match" model of allopolyploidy, in which subgenome origin drives homoeolog loss propensities but where genes from different subgenomes function together without difficulty.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Dark energy survey year 3 results: likelihood-free, simulation-based w CDM inference with neural compression of weak-lensing map statistics

We present simulation-based cosmological wcold dark matter (wCDM) inference using dark energy survey year 3 weak-lensing maps, via neural data compression of weak-lensing map summary statistics: power spectra, peak counts, and direct map-level compression/inference with convolutional neural networks (CNN). Using simulation-based inference, also known as likelihood-free or implicit inference, we use forward-modelled mock data to estimate posterior probability distributions of unknown parameters. This approach allows all statistical assumptions and uncertainties to be propagated through the forward-modelled mock data; these include sky masks, non-Gaussian shape noise, shape measurement bias, source galaxy clustering, photometric redshift uncertainty, intrinsic galaxy alignments, non-Gaussian density fields, neutrinos, and non-linear summary statistics. We include a series of tests to validate our inference results. This paper also describes the Gower Street simulation suite: 791 full-sky pkdgrav3 dark matter simulations, with cosmological model parameters sampled with a mixed active-learning strategy, from which we construct over 3000 mock dark energy survey lensing data sets. For wCDM inference, for which we allow –1 < w < –$\frac{1}{3}$⁠, our most constraining result uses power spectra combined with map-level (CNN) inference. Using gravitational lensing data only, this map-level combination gives Ω m = 0.283$^{+0.020}_{–0.027}$⁠, S 8 = 0.804$^{+0.025}_{–0.017⁠}$, and w < –0.80 (with a 68 per cent credible interval); compared to the power spectrum inference, this is more than a factor of two improvement in dark energy parameter (Ω⁠ DE , w⁠) precision.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

Scalable Bayesian Nonparametric Clustering and Classification

We develop a scalable multi-step Monte Carlo algorithm for inference under a large class of nonparametric Bayesian models for clustering and classification. Each step is “embarrassingly parallel” and can be implemented using the same Markov chain Monte Carlo sampler. The simplicity and generality of our approach makes inference for a wide range of Bayesian nonparametric mixture models applicable to large datasets. Specifically, we apply the approach to inference under a product partition model with regression on covariates. We show results for inference with two motivating data sets: a large set of electronic health records (EHR) and a bank telemarketing dataset. We find interesting clusters and competitive classification performance relative to other widely used competing classifiers. Supplementary materials for this article are available online.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Observed and Simulated Sensitivities of Spring Greenup to Preseason Climate in Northern Temperate and Boreal Regions: Climate control on spring phenology

Vegetation phenology plays an important role in regulating land-atmosphere energy, water, and trace-gas exchanges. Changes in spring greenup (SG) have been documented in the past half-century in response to ongoing climate change. We use normalized difference vegetation index generated from NOAA's advanced very high resolution radiometer data in the Global Inventory Modeling and Monitoring Study project over the 1982–2005 period, coupled with climate reanalysis (Climate Research Unit-National Centers for Environmental Prediction) to investigate the SG responses to preseason climate change in northern temperate and boreal regions. We compared these observed responses to the simulated SG responses to preseason climate inferred from the Earth system models (ESMs) participating in the Coupled Model Intercomparison Project Phase 5 (CMIP5) over 1982–2005. The observationally inferred SG suggests that there has been an advance of about 1 days per decade between 1982 and 2005 in the northern midlatitude to high latitude, with significant spatial heterogeneity. The spatial heterogeneity of the SG advance results from heterogeneity in the change of the preseason climate as well as varied vegetation responses to the preseason climate across biomes. The SG to preseason temperature sensitivity is highest in forests other than deciduous needleleaf forests, followed by temperate grasslands and woody savannas. The SG in deciduous needleleaf forests, open shrublands, and tundra is relatively insensitive to preseason temperature. Although the extent of regions where the SG is sensitive to preseason precipitation is smaller than the extent of regions where the SG is sensitive to preseason temperature, the biomes that are more sensitive to temperature are also more sensitive to precipitation, suggesting the interactive control of temperature and precipitation. Finally, in the mean, the CMIP5 ESMs reproduced the dominant latitudinal preseason climate trends and SG advances. However, large biases in individual ESMs for the preseason period, climate, and SG sensitivity imply needed model improvements to climate prediction and phenological process parameterizations.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2019

Multi-region Modeling of Daily Global Solar Radiation with Artificial Intelligence Ensemble

Solar radiation data are crucial for the design and evaluation of solar energy systems, climatic studies, water resources management, estimating crop productivity, etc. As so, for locations where direct measurements are not available, reliable models may be developed to estimate solar radiation from more readily available data. In this study, two artificial intelligence (AI)-based models including artificial neural network and adaptive neuro-fuzzy inference systems, three temperature-based empirical models including Meza–Varas, Hargreaves–Samani, and Chen, and a conventional multi-linear regression (MLR) model were employed for multi-region daily global solar radiation estimation for Iraq. To ensure appropriate selection of input variables, sensitivity analysis was conducted to determine the dominant parameters. Finally, two ensemble approaches, neural average ensemble and simple average ensemble, were applied to improve the performance of the single models. For this purpose, daily meteorological data of maximum temperature (T{sub max}), minimum temperature (T{sub min}), mean temperature (T{sub mean}), relative humidity (R{sub H}), and wind speed (U{sub 2}) were obtained from January 2006 to December 2016 from four major cities in Iraq representing, north, west, south, and east regions. The results revealed that temperatures (T{sub max}, T{sub mean}, T{sub min}) and relative humidity are the dominant parameters. While temperature-based empirical models and MLR model could be employed to achieve reliable results, AI-based models are superior in performance to other models. Also promising improvement in daily global solar radiation modeling could be achieved by model ensemble. The results of this study affirmed that the provided ensemble approaches can increase the performance of single models up to 19.19%, 7.59%, and 16.81% in training, validation, and testing steps, respectively.

58 GEOSCIENCES↗
DOE OSTI2021

Evaluation of silicon carbide as a divertor armor material in DIII-D H-mode discharges

Silicon carbide (SiC) represents a promising but largely untested plasma-facing material (PFM) for next-step fusion devices. In this work, an analytic mixed-material erosion model is developed by calculating the physical (via SDTrimSP) and chemical (via empirical scalings) sputtering yield from SiC, Si, and C. The Si content in the near-surface SiC layer is predicted to increase during D plasma bombardment due to more efficient physical and chemical sputtering of C relative to Si. Silicon erosion from SiC thereby occurs primarily from sputtering of the enriched Si layer, rather than directly from the SiC itself. SiC coatings on ATJ graphite, manufactured via chemical vapor deposition, were exposed to repeated H-mode plasma discharges in the DIII-D tokamak to test this model. The qualitative trends from analytic modeling are reproduced by the experimental measurements, obtained via spectroscopic inference using the S/XB method. Quantitatively the model slightly under-predicts measured erosion rates, which is attributed to uncertainties in the ion impact angle distribution, as well as the effect of edge-localized modes. After exposure, minimal changes to the macroscopic or microscopic surface morphology of the SiC coatings were observed. Compositional analysis reveals Si enrichment of about 10%, in line with expectations from the erosion model. Extrapolating to a DEMO-type device, an order-of-magnitude decrease in impurity sourcing, and up to a factor of 2 decrease in impurity radiation, is expected with SiC walls, relative to graphite, if low C plasma impurity content can be achieved. Furthermore, these favorable erosion properties motivate further investigations of SiC as a low-Z, non-metallic PFM.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2017

The Kinematics of the Permitted C ii λ 6578 Line in a Large Sample of Planetary Nebulae

We present spectroscopic observations of the C ii λ 6578 permitted line for 83 lines of sight in 76 planetary nebulae at high spectral resolution, most of them obtained with the Manchester Echelle Spectrograph on the 2.1 m telescope at the Observatorio Astronómico Nacional on the Sierra San Pedro Mártir. We study the kinematics of the C ii λ 6578 permitted line with respect to other permitted and collisionally excited lines. Statistically, we find that the kinematics of the C ii λ 6578 line are not those expected if this line arises from the recombination of C{sup 2+} ions or the fluorescence of C{sup +} ions in ionization equilibrium in a chemically homogeneous nebular plasma, but instead its kinematics are those appropriate for a volume more internal than expected. The planetary nebulae in this sample have well-defined morphology and are restricted to a limited range in H α line widths (no large values) compared to their counterparts in the Milky Way bulge; both these features could be interpreted as the result of young nebular shells, an inference that is also supported by nebular modeling. Concerning the long-standing discrepancy between chemical abundances inferred from permitted and collisionally excited emission lines in photoionized nebulae, our results imply that multiple plasma components occur commonly in planetary nebulae.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2017

Stan : A Probabilistic Programming Language

Stan is a probabilistic programming language for specifying statistical models. A Stan program imperatively defines a log probability function over parameters conditioned on specified data and constants. As of version 2.14.0, Stan provides full Bayesian inference for continuous-variable models through Markov chain Monte Carlo methods such as the No-U-Turn sampler, an adaptive form of Hamiltonian Monte Carlo sampling. Penalized maximum likelihood estimates are calculated using optimization methods such as the limited memory Broyden-Fletcher-Goldfarb-Shanno algorithm. Stan is also a platform for computing log densities and their gradients and Hessians, which can be used in alternative algorithms such as variational Bayes, expectation propagation, and marginal inference using approximate integration. To this end, Stan is set up so that the densities, gradients, and Hessians, along with intermediate quantities of the algorithm such as acceptance probabilities, are easily accessible. Stan can also be called from the command line using the cmdstan package, through R using the rstan package, and through Python using the pystan package. All three interfaces support sampling and optimization-based inference with diagnostics and posterior analysis. rstan and pystan also provide access to log probabilities, gradients, Hessians, parameter transforms, and specialized plotting.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Mitigation of spatial nonstationarity with vision transformers

Spatial nonstationarity, the location variance of features’ statistical distributions, is ubiquitous in many natural settings. For example, in geological reservoirs rock matrix porosity varies vertically due to geomechanical compaction trends, in mineral deposits grades vary due to sedimentation and concentration processes, in hydrology rainfall varies due to the atmosphere and topography interactions, and in metallurgy crystalline structures vary due to differential cooling. Conventional geostatistical modeling workflows rely on the assumption of stationarity to be able to model spatial features for geostatistical inference. Nevertheless, this is often not a realistic assumption when dealing with nonstationary spatial data and this has motivated a variety of nonstationary spatial modeling workflows such as trend and residual decomposition, cosimulation with secondary features, and spatial segmentation and independent modeling over stationary subdomains. The advent of deep learning technologies has enabled new workflows for modeling spatial relationships. However, there is a paucity of demonstrated best practice and general guidance on mitigation of spatial nonstationarity with deep learning in the geospatial context. We demonstrate the impact of two common types of geostatistical spatial nonstationarity on deep learning model prediction performance and propose the mitigation of such impacts using self-attention (vision transformer) models. We demonstrate the utility of vision transformers for the mitigation of nonstationarity with relative errors as low as 10%, exceeding the performance of alternative deep learning methods such as convolutional neural networks. We establish best practice by demonstrating the ability of self-attention networks for modeling large-scale spatial relationships in the presence of commonly observed geospatial nonstationarity.

58 GEOSCIENCES↗
DOE OSTI2018

Microstructural evolution of an aluminide coating on alloy 625 during wet air exposure at 900 °C and 1000 °C

The microstructural changes of the aluminized alloy 625 during cyclic oxidation in air + 6% H 2 O at 900 °C and 1000 °C were analyzed using optical metallography (OM), scanning electron microscopy (SEM) with energy and wave length dispersive X-ray analysis (EDX/WDX) as well as electron backscatter diffraction (EBSD). An in-house developed thermodynamic-kinetic procedure was employed to predict the microstructural evolution of aluminized alloy 625 during high temperature exposure by considering simultaneously occurring surface oxidation and interdiffusion processes. Because of the lack of mobility data for the relevant alloying elements in the σ-phase, assumptions for the mobilities were made based on the value of the mobilities in α-Cr. Despite these assumptions, the calculated results were found to be in good agreement with experimental observations. The complete depletion of β-NiAl in the coating observed during exposure at 1000 °C was correctly predicted by the model. The model was also able to infer dissolution of the precipitate phases α-Cr and σ in the interdiffusion zone during exposures at 900 °C and 1000 °C. The model was yet unable to predict the formation of the μ-phase in the alloy after 1000 h of exposure at 1000 °C. The developed modelling method offers the potential to predict microstructural changes of aluminized nickel base alloys thus reducing cost and time consuming experimental efforts.

36 MATERIALS SCIENCE↗
DOE OSTI2021

On the Stochastic Stability of Deep Markov Models

Deep Markov models (DMM) are generative models which are scalable and expressive generalization of Markov models for representation, learning, and inference problems. DMMs using deep neural networks to parametrize the transition of Markov probability distributions have recently been shown to provide more expressiveness in modeling sequential data and dynamical system responses. However, the fundamental stochastic stability guarantees of such models have not been thoroughly investigated. In this paper, we present a rigorous analytical method to prove the necessary and sufficient conditions of DMM's stochastic stability. This task is achieved by spectral analysis of the efficiently computed Jacobians of probabilistic maps modeled by deep neural networks. We make theoretical connections between the eigenvalues of neural network's weights and the different activation function types used on the stability and overall dynamic behavior of DMMs with Gaussian distributions. We empirically substantiate our theoretical results on stochastic stability and eigenvalue spectra via several numerical experiments. Formal stability guarantees of DMMs can substantially improve their robustness and trustworthiness, necessary for reliable use in safety-critical real-world applications.

Drgona, Jan↗
DOE OSTI2019

Approximate inference for constructing astronomical catalogs from images

Here, we present a new, fully generative model for constructing astronomical catalogs from optical telescope image sets. Each pixel intensity is treated as a random variable with parameters that depend on the latent properties of stars and galaxies. These latent properties are themselves modeled as random. We compare two procedures for posterior inference. One procedure is based on Markov chain Monte Carlo (MCMC) while the other is based on variational inference (VI). The MCMC procedure excels at quantifying uncertainty, while the VI procedure is 1000 times faster. On a supercomputer, the VI procedure efficiently uses 665,000 CPU cores to construct an astronomical catalog from 50 terabytes of images in 14.6 minutes, demonstrating the scaling characteristics necessary to construct catalogs for upcoming astronomical surveys.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Canonical and Noncanonical Hamiltonian Operator Inference

SAND2024-08577O Canonical and Noncanonical Hamiltonian Operator Inference is research code that applies old and new model reduction algorithms to well-known and publicly available "toy" problems. It applies operator inference (OpInf) methods for model reduction to several benchmark hyperbolic partial differential equations: linear wave, Korteweg-de Vries, Benjamin-Bona-Mahoney, and linear elasticity. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

SciDAC↗
DOE OSTI2021

Structurally flexible cloud microphysics, observationally constrained at all scales via ML-accelerated Bayesian inference

We discuss the challenge of developing observationally informed parameterizations of microphysics for use at a hierarchy of modeling scales. Our proposed approach is applicable to any domain that suffers from a two-fold parameterization problem, where physical processes are not resolved at the model scale (the first problem), and those processes are uncertain at any scale (the second problem). For such problems, a physical approach facilitates modeling across scales, as well as systematic observational inference accelerated by machine learning (ML) surrogate models, and quantification of physical uncertainties.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

The Importance of Being Adaptable: An Exploration of the Power and Limitations of Domain Adaptation for Simulation-Based Inference with Galaxy Clusters

The application of deep machine learning methods in astronomy has exploded in the last decade, with new models showing remarkably improved performance on benchmark tasks. Not nearly enough attention is given to understanding the models' robustness, especially when the test data are systematically different from the training data, or "out of domain." Domain shift poses a significant challenge for simulation-based inference, where models are trained on simulated data but applied to real observational data. In this paper, we explore domain shift and test domain adaptation methods for a specific scientific case: simulation-based inference for estimating galaxy cluster masses from X-ray profiles. We build datasets to mimic simulation-based inference: a training set from the Magneticum simulation, a scatter-augmented training set to capture uncertainties in scaling relations, and a test set derived from the IllustrisTNG simulation. We demonstrate that the Test Set is out of domain in subtle ways that would be difficult to detect without careful analysis. We apply three deep learning methods: a standard neural network (NN), a neural network trained on the scatter-augmented input catalogs, and a Deep Reconstruction-Regression Network (DRRN), a semi-supervised deep model engineered to address domain shift. Although the NN improves results by 17% in the Training Data, it performs 40% worse on the out-of-domain Test Set. Surprisingly, the Scatter-Augmented Neural Network (SANN) performs similarly. While the DRRN is successful in mapping the training and Test Data onto the same latent space, it consistently underperforms compared to a straightforward Yx scaling relation. These results serve as a warning that simulation-based inference must be handled with extreme care, as subtle differences between training simulations and observational data can lead to unforeseen biases creeping into the results.

Ntampaka, Michelle [Baltimore, Space Telescope Sci↗
DOE OSTI2017

A Bayesian approach to modeling diffraction profiles and application to ferroelectric materials

A new statistical approach for modeling diffraction profiles is introduced, using Bayesian inference and a Markov chain Monte Carlo (MCMC) algorithm. This method is demonstrated by modeling the degenerate reflections during application of an electric field to two different ferroelectric materials: thin-film lead zirconate titanate (PZT) of composition PbZr 0.3 Ti 0.7 O 3 and a bulk commercial PZT polycrystalline ferroelectric. Here, the new method offers a unique uncertainty quantification of the model parameters that can be readily propagated into new calculated parameters.

36 MATERIALS SCIENCE↗