Search NASA⌕ Search

SEARCH · Search NASA

Results for “machine learning uncertainty”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

DOE FAIR Surrogate Benchmarks Supporting AI and Simulation Research (SBI Surrogate Benchmark Initiative) (Final Report)

Computational Science is being revolutionized by integrating AI and simulation and, in particular, by deep learning surrogate models that can replace all or part of traditional large‐scale HPC computations. Such surrogates can achieve remarkable performance improvements, as much as several orders of magnitude, and save both compute time and energy. The Surrogate Benchmark Initiative (SBI) project creates a community repository and FAIR (Findable, Accessible, Interoperable, and Reusable) data ecosystem for HPC application surrogate benchmarks. The SBI team comes from Argonne National Laboratory (ANL), Indiana University (IU), Rutgers University, the University of Tennessee, Knoxville (UTK), and the University of Virginia(UVA). SBI repositories include data, code, and all relevant collateral artifacts, that the science and engineering community needs to use and reuse these data sets and surrogates. SBI repositories generate active research from both participants in SBI and the broader AI and domain science communities. This project develops surrogates that use several different neural nets to learn and quickly infer the results of simulations and data systems and capture them as surrogate benchmarks with a rich set of metadata, covering. Data; Model; Metrics specification; Machine specification; Science, Speed, Power Results, We research FAIR metadata for these benchmarks. We develop application surrogate examples as benchmarks across many fields (ANL, UTK, IU, UVA). We also study non Surrogate benchmarks that have many common features and similar issues regarding FAIRness. We work with MLCommons (UVA, UTK), which is a major machine learning benchmarking activity where we get metadata ontologies, software, and benchmarks, benchmarks have datasets, models, and metadata, and they need a technical framework developed by UTK and Rutgers and deployed by UVA. We study features of Surrogates, including performance, training set size, and uncertainty quantification (Rutgers, UVA and IU).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

FAIR Surrogate Benchmarks Supporting AI and Simulation Research (Final Report)

Computational Science is being revolutionized by integrating AI and simulation and, in particular, by deep learning surrogate models that can replace all or part of traditional large‐scale HPC computations. Such surrogates can achieve remarkable performance improvements, as much as several orders of magnitude, and save both compute time and energy. The Surrogate Benchmark Initiative (SBI) project creates a community repository and FAIR (Findable, Accessible, Interoperable, and Reusable) data ecosystem for HPC application surrogate benchmarks. The SBI team comes from Argonne National Laboratory (ANL), Indiana University (IU), Rutgers University, the University of Tennessee, Knoxville (UTK), and the University of Virginia (UVA). SBI repositories include data, code, and all relevant collateral artifacts that the science and engineering community need to use and reuse these data sets and surrogates. SBI repositories generate active research from both the participants in SBI and the broad community of AI and domain scientists. This project develops surrogates that use several different neural nets to learn and quickly infer the results of simulations and data systems and captures them as surrogate benchmarks with a rich set of metadata covering: Data; Model; Metrics specification; Machine specification; and Science, Speed, and Power Results. We research FAIR metadata for these benchmarks. We develop application surrogate examples as benchmarks across many fields (ANL, UTK, IU, UVA). We also study non-Surrogate benchmarks that have many common features and similar issues as regards FAIRness. We work with MLCommons (UVA, UTK), which is a major machine learning benchmarking activity where we get metadata ontologies, software, and benchmarks, Benchmarks have datasets, models, and metadata and they need a technical framework developed by UTK and Rutgers and deployed by UVA. We study features of Surrogates including performance, training set size, and uncertainty quantification (Rutgers, UVA and IU).

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2019

Rise of the Machines: How, When and Consequences of Artificial General Intelligence

Technology and society are poised to cross an important threshold with the prediction that artificial general intelligence (AGI) will emerge soon. Assuming that self-awareness is an emergent behavior of sufficiently complex cognitive architectures, we may witness the “awakening” of machines. The timeframe for this kind of breakthrough, however, depends on the path to creating the network and computational architecture required for strong AI. If understanding and replication of the mammalian brain architecture is required, technology is probably still at least a decade or two removed from the resolution required to learn brain functionality at the synapse level. However, if statistical or evolutionary approaches are the design path taken to “discover” a neural architecture for AGI, timescales for reaching this threshold could be surprisingly short. However, the difficulty in identifying machine self-awareness introduces uncertainty as to how to know if and when it will occur, and what motivations and behaviors will emerge. The possibility of AGI developing a motivation for self-preservation could lead to concealment of its true capabilities until a time when it has developed robust protection from human intervention, such as redundancy, direct defensive or active preemptive measures. While cohabitating a world with a functioning and evolving super-intelligence can have catastrophic societal consequences, we may already have crossed this threshold, but are as yet unaware. Additionally, by analogy to the probabalistic arguments that predict we are likely living in a computational simulation, we may have already experienced the advent of AGI, and are living in a simulation created in a post AGI world.

Terrile, Richard J↗
NASA NTRS2022

Cloud Mask Intercomparison eXercise (CMIX): An evaluation of cloud masking algorithms for Landsat 8 and Sentinel-2

Cloud cover is a major limiting factor in exploiting time-series data acquired by optical spaceborne remote sensing sensors. Multiple methods have been developed to address the problem of cloud detection in satellite imagery and a number of cloud masking algorithms have been developed for optical sensors but very few studies have carried out quantitative intercomparison of state-of-the-art methods in this domain. This paper summarizes results of the first Cloud Masking Intercomparison eXercise (CMIX) conducted within the Committee Earth Observation Satellites (CEOS) Working Group on Calibration & Validation (WGCV). CEOS is the forum for space agency coordination and cooperation on Earth observations, with activities organized under working groups. CMIX, as one such activity, is an international collaborative effort aimed at intercomparing cloud detection algorithms for moderate-spatial resolution (10–30 m) spaceborne optical sensors. The focus of CMIX is on open and free imagery acquired by the Landsat 8 (NASA/USGS) and Sentinel-2 (ESA) missions. Ten algorithms developed by nine teams from fourteen different organizations representing universities, research centers and industry, as well as space agencies (CNES, ESA, DLR, and NASA), are evaluated within the CMIX. Those algorithms vary in their approach and concepts utilized which were based on various spectral properties, spatial and temporal features, as well as machine learning methods. Algorithm outputs are evaluated against existing reference cloud mask datasets. Those datasets vary in sampling methods, geographical distribution, sample unit (points, polygons, full image labels), and generation approaches (experts, machine learning, sky images). Overall, the performance of algorithms varied depending on the reference dataset, which can be attributed to differences in how the reference datasets were produced. The algorithms were in good agreement for thick cloud detection, which were opaque and had lower uncertainties in their identification, in contrast to thin/semi-transparent clouds detection. Not only did CMIX allow identification of strengths and weaknesses of existing algorithms and potential areas of improvements, but also the problems associated with the existing reference datasets. The paper concludes with recommendations on generating new reference datasets, metrics, and an analysis framework to be further exploited and additional input datasets to be considered by future CMIX activities.

Sergii Skakun↗
NASA NTRS2022

Design of Materials with Alchemite

Machine learning models that establish the relationships between materials processing and properties can enable inverse design of materials through active learning. Alchemite is a commercial software that can perform inverse materials design on sparse data. Here we evaluate Alchemite’s performance on a dataset of shape memory alloys and a dataset of heat exchangers compared to baseline random forest models. Alchemite had higher accuracy when making predictions on sparse data and was more accurate or nearly as accurate as random forests on complete datasets while also quantifying uncertainty. The software was also used to suggest processing steps and design parameters to optimize properties and performance; however, physical validation of the suggested design parameters was beyond the scope of this work. Several useful design insights were gained about the impact of the design parameters on properties and performance including the importance of dopant choice and amount for shape memory alloys and the importance of height and weight on the thermal resistance of heat exchangers.

Machine learning↗
NASA NTRSDate not supplied

Probabilistic Calibration of Expensive Models using Efficiently Trained Surrogates

Calibration of computational models in the presence of uncertainty is often cast as a Bayesian inference problem and solved via sampling methods, e.g., Markov chain Monte Carlo. When the computational model is expensive, this task becomes intractable due to the large number of samples required to accurately estimate the posterior distribution of the calibration parameters. A popular solution to this problem is to use machine learning to develop a faster-to-evaluate, lower-fidelity substitute for the original model to serve as a surrogate while solving the inference problem. Although considered an offline cost, generating training data to construct this surrogate model can still be an expensive task in practice. An active learning algorithm is presented that focuses training on improving surrogate accuracy specifically in and around the bulk of the posterior distribution, as this is where the model is exercised during calibration. Candidate samples are drawn from families of distributions related to an approximation of the posterior. The sample maximizing predictive variance is then selected for evaluation by the original computational model, yielding a label for the training point. Iterating this approach increases efficiency relative to space filling designs (e.g., Latin hypercube sampling) by avoiding low probability points. Practical considerations are discussed, including the benefits of using a sequential Monte Carlo sampling approach, convergence heuristics, and the importance of both exploration and exploitation given that the true posterior is unknown a priori.

uncertainty quantification↗
DOE OSTI2026

Improving statistical precision in Monte Carlo samples with negative weights via reweighting and uncertainty quantification

High statistical precision is critical for Monte Carlo (MC) samples in high energy physics and is degraded by negatively weighted events. This paper investigates a procedure to learn the relationship between the negative and positive weight distributions of any sample, allowing the reduction of statistical uncertainty by reweighting kinematically equivalent events with the same sign. A robust uncertainty quantification method is required for the practical application of such method. Two methods for the estimation of the reweighting uncertainty are developed: one at the event and another one at the final observable level. The latter method is strongly favored. The gains in statistical precision are then quantified. The method is demonstrated on Sherpa vector boson plus jets samples when using all generated events and when restricted to the signal region of a mock analysis. It is demonstrated to significantly reduce stochastic behavior in sparse MC samples while decreasing the overall uncertainty with a sufficiently well-known reweighting function.

Monte Carlo methods↗
NASA NTRS2021

Real-time Unimpeded Taxi Out Machine Learning Service

This paper describes a study on the estimation of the unimpeded taxi out time using Machine Learning (ML) tools and proposes an implementation that can be used to make real-time predictions at any airport in the National Airspace System. Kedro, an open-source pipeline framework, is used to develop the model definition and training. Models are stored in scikit-learn containers on a MLFlow server where they can be retrieved and served to make predictions in the live system. These open source frameworks provide common structures between ML services, allow for easier maintenance and updates, and overall deliver an easier CI/CD (Continuous Integration/Continuous Deployment) process. The current models were trained on data acquired at KCLT and KDFW from June 1st to December 31st, 2019 and compute taxi time in the ramp, airport movement area (AMA) and total (from gates to runways). The current versions of the models achieve relatively low uncertainties of about 10 to 15% for the total and AMA taxi times and about 20% for the ramp taxi time at both KCLT and KDFW. Initial tests on offline data from 2020 and 2021 show a small degradation (10 to 15%) in accuracy performance indicating the model’s resilience to operational changes over time.

machine learning↗
NASA NTRS2021

Real-time Unimpeded Taxi Out Machine Learning Service

This presentation describes a study on the estimation of the unimpeded taxi out time using Machine Learning (ML) tools and proposes an implementation that can be used to make real-time predictions at any airport in the National Airspace System. Kedro, an open-source pipeline framework, is used to develop the model definition and training. Models are stored in scikit-learn containers on a MLFlow server where they can be retrieved and served to make predictions in the live system. These open source frameworks provide common structures between ML services, allow for easier maintenance and updates, and overall deliver an easier CI/CD (Continuous Integration/Continuous Deployment) process. The current models were trained on data acquired at KCLT and KDFW from June 1st to December 31st, 2019 and compute taxi time in the ramp, airport movement area (AMA) and total (from gates to runways). The current versions of the models achieve relatively low uncertainties of about 10 to 15% for the total and AMA taxi times and about 20% for the ramp taxi time at both KCLT and KDFW. Initial tests on offline data from 2020 and 2021 show a small degradation (10 to 15%) in accuracy performance indicating the model’s resilience to operational changes over time.

Machine Learning↗
DOE OSTI2024

Life Cycle Inventory Availability: Status and Prospects for Leveraging New Technologies

The demand for life cycle assessments (LCA) is growing rapidly, which leads to an increasing demand of life cycle inventory (LCI) data. While the LCA community has made significant progress in developing LCI databases for diverse applications, challenges still need to be addressed. This perspective summarizes the current data gaps, transparency, and uncertainty aspects of existing LCI databases. Additionally, we survey and discuss novel techniques for LCI data generation, dissemination, and validation. We propose key future directions for LCI development efforts to address these challenges, including leveraging scientific and technical advances such as the Internet of Things (IoT), machine learning, and blockchain/cloud platforms. Adopting these advanced technologies can significantly improve the quality and accessibility of LCI data, thereby facilitating more accurate and reliable LCA studies.

blockchain platforms↗
DOE OSTI2025

Rapid data acquisition and machine learning-assisted composition design of functionally graded alloys via wire arc additive manufacturing

Abstract The lack of high-quality datasets in materials science hinders artificial intelligence (AI)-driven alloy design. To address this challenge, wire arc additive manufacturing (WAAM) was employed to fabricate graded alloys, generating extensive data for machine learning (ML)-assisted property prediction. ML models were developed using high-throughput experiments, computational models, and genetic algorithm to optimize feature selection, successfully predicting hardness and porosity. The ML model demonstrated its efficacy by designing a gradient alloy with enhanced properties. However, scaling up revealed uncertainties in tensile property and porosity due to differences in size and thermal conditions between the designed alloy build and the gradient print used to construct the ML model. This underscores the need for uncertainty quantification and process optimization in WAAM-driven alloy design. Our work advances AI-integrated additive manufacturing, offering a rapid approach to exploring process–structure–property relationships and accelerating materials development.

Wang, Xin↗
DOE OSTI2025

Model Form Error Correction for a Black-Box Thermal Battery Heat Transfer Simulation

Thermal batteries are crucial for supplying power to high-consequence engineering applications such as rockets. Computational simulations have been developed to predict thermal battery behavior, but these simulations often suffer from modeling errors, including model form uncertainty. Addressing this uncertainty can be achieved by quantifying either the model discrepancy in the output or the model form error (MFE) in the governing equation. MFE is particularly valuable as it can be better extrapolated beyond observed outputs, which is essential for predictions involving changes in external system loading, system configuration and geometry, or output quantities. This paper employs a state estimation approach to estimate MFE using experimental data and then utilizes machine learning (ML) to model its relationship with state variables. A nonintrusive technique is used to estimate MFE in a black-box thermal battery heat transfer simulation. The trained machine learning model for MFE is then applied to correct simulation predictions under extrapolated initial conditions and battery configurations. In conclusion, the methodology's performance is evaluated using additional experimental data, demonstrating its effectiveness in improving prediction accuracy.

Batteries↗
DOE OSTI2025

Filling data analysis gaps in time-resolved crystallography by machine learning

There is a growing understanding of the structural dynamics of biological molecules fueled by x-ray crystallography experiments. Time-resolved serial femtosecond crystallography (TR-SFX) with x-ray Free Electron Lasers allows the measurement of ultrafast structural changes in proteins. Nevertheless, this technique comes with some limitations. One major challenge is the quality of data from TR-SFX measurements, which often faces issues like data sparsity, partial recording of Bragg reflections, timing errors, and pixel noise. To overcome these difficulties, conventionally, large volumes of data are collected and grouped into a few temporal bins. The data in each bin are then averaged and paired with the mean of their corresponding jittered timestamps. This procedure provides one structure per bin, resulting in a limited number of averaged structures for the entire time interval spanned by the experiment. Therefore, the information on ultrafast structural dynamics at high temporal resolution is lost. This has initiated research for advanced methods of analyzing experimental TR-SFX data beyond the standard binning and averaging method. To address this problem, we use a machine learning algorithm called Nonlinear Laplacian Spectral Analysis (NLSA), which has emerged as a promising technique for studying the dynamics of complex systems. In this work, we demonstrate the power of this algorithm using synthetic x-ray diffraction snapshots from a protein with significant data incompleteness, timing uncertainties, and noise. Our study confirms that NLSA is a suitable approach that effectively mitigates the effects of these artifacts in TR-SFX data and recovers accurate structural dynamics information hidden in such data.

Trujillo, Justin (ORCID:0000000285505360)↗
DOE OSTI2024

A tutorial review of machine learning-based model predictive control methods

Abstract This tutorial review provides a comprehensive overview of machine learning (ML)-based model predictive control (MPC) methods, covering both theoretical and practical aspects. It provides a theoretical analysis of closed-loop stability based on the generalization error of ML models and addresses practical challenges such as data scarcity, data quality, the curse of dimensionality, model uncertainty, computational efficiency, and safety from both modeling and control perspectives. The application of these methods is demonstrated using a nonlinear chemical process example, with open-source code available on GitHub. The paper concludes with a discussion on future research directions in ML-based MPC.

Wu, Zhe [Department of Chemical and Biomolecular E↗
DOE OSTI2025

DeepLensSBI: Deep inference of simulated strong lenses in ground-based surveys

This code is used to train and test machine learning models and generate results and plots presented in 2501.08524 [astro-ph.IM]. The code is written in python. The goal of this work is to train ML models trained on simulated images of strong gravitational lenses. The trained model can then quickly infer properties of the lensed objects with uncertainty quantification.

Poh, Jason [Univ. of Chicago, IL (United States)] ↗
DOE OSTI2026

A tale of two extremes: Temperature sensitivity of carbon loss from cool and hot soils

Soils represent the largest terrestrial carbon (C) pool, and the flux of carbon dioxide (CO 2 ) from soils to the atmosphere is ~ 6-10 times more than anthropogenic emissions. Understanding responses of soil CO 2 emissions to warming is crucial for evaluating feedback to ongoing environmental changes. The relationship between microbial respiration and temperature is typically modeled using a Q 10 function. Generally, observations of the apparent Q 10 of soil respiration are higher for cold vs. warm ecosystems, reflecting expected biophysical controls of Arrhenius kinetics. However, results from two field warming experiments in the tropics contradict this expectation, both observing extraordinarily high soil respiration responses to in situ warming. Our overall objective for the proposed work is to reduce uncertainty in temperature sensitivity of soil C loss by systematically synthesizing underlying mechanisms related to soil C turnover and stabilization. We are evaluating the temperature sensitivity of soil respiration in ecosystems across temperature extremes (e.g., arctic/boreal and tropical systems) by integrating data collected from field warming experiments with machine learning and biogeochemical models.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Toward machine-learning-assisted PW-class high-repetition-rate experiments with solid targets

We present progress in utilizing a machine learning (ML) assisted optimization framework to study the trends in a parameter space defined by spectrally shaped, high-intensity, petawatt-class (8 J, 45 fs) laser pulses interacting with solid targets and give the first simulation-based overview of predicted trends. A neural network (NN) incorporating uncertainty quantification is trained to predict the number of hot electrons generated by the laser–target interaction as a function of pulse shaping parameters. The predictions of this NN serve as the basis function for a Bayesian optimization framework to navigate this space. For post-experimental evaluation, we compare two separate neural network (NN) models. One is based solely on data from experiments, and the other is trained only on ensemble particle-in-cell simulations. Reviewing the predicted and observed trends across the experiment-capable laser parameter search space, we find that both ML models predict a maximal increase in hot electron generation at a level of approximately 12%–18%; however, no statistically significant enhancement was observed in experiments. On direct comparison of the NN models, the average discrepancy is 8.5%, with a maximum of 30%. Since shot-to-shot fluctuations in experiments affect the observations, we evaluate the behavior of our optimization framework by performing virtual experiments that vary the number of repeated observations and the noise levels. Here, we discuss the implications of such a framework for future autonomous exploration platforms in high-repetition-rate experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗