Search NASA⌕ Search

SEARCH · Search NASA

Results for “Scientific Machine Learning”

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 775 records · Page 43
All sourcesNASA NTRSDOE OSTI
DOE OSTI2010

Statistical Exploration of Electronic Structure of Molecules from Quantum Monte-Carlo Simulations

In this report, we present results from analysis of Quantum Monte Carlo (QMC) simulation data with the goal of determining internal structure of a 3N-dimensional phase space of an N-electron molecule. We are interested in mining the simulation data for patterns that might be indicative of the bond rearrangement as molecules change electronic states. We examined simulation output that tracks the positions of two coupled electrons in the singlet and triplet states of an H2 molecule. The electrons trace out a trajectory, which was analyzed with a number of statistical techniques. This project was intended to address the following scientific questions: (1) Do high-dimensional phase spaces characterizing electronic structure of molecules tend to cluster in any natural way? Do we see a change in clustering patterns as we explore different electronic states of the same molecule? (2) Since it is hard to understand the high-dimensional space of trajectories, can we project these trajectories to a lower dimensional subspace to gain a better understanding of patterns? (3) Do trajectories inherently lie in a lower-dimensional manifold? Can we recover that manifold? After extensive statistical analysis, we are now in a better position to respond to these questions. (1) We definitely see clustering patterns, and differences between the H2 and H2tri datasets. These are revealed by the pamk method in a fairly reliable manner and can potentially be used to distinguish bonded and non-bonded systems and get insight into the nature of bonding. (2) Projecting to a lower dimensional subspace ({approx}4-5) using PCA or Kernel PCA reveals interesting patterns in the distribution of scalar values, which can be related to the existing descriptors of electronic structure of molecules. Also, these results can be immediately used to develop robust tools for analysis of noisy data obtained during QMC simulations (3) All dimensionality reduction and estimation techniques that we tried seem to indicate that one needs 4 or 5 components to account for most of the variance in the data, hence this 5D dataset does not necessarily lie on a well-defined, low dimensional manifold. In terms of specific clustering techniques, K-means was generally useful in exploring the dataset. The partition around medoids (pam) technique produced the most definitive results for our data showing distinctive patterns for both a sample of the complete data and time-series. The gap statistic with tibshirani criteria did not provide any distinction across the 2 dataset. The gap statistic w/DandF criteria, Model based clustering and hierarchical modeling simply failed to run on our datasets. Thankfully, the vanilla PCA technique was successful in handling our entire dataset. PCA revealed some interesting patterns for the scalar value distribution. Kernel PCA techniques (vanilladot, RBF, Polynomial) and MDS failed to run on the entire dataset, or even a significant fraction of the dataset, and we resorted to creating an explicit feature map followed by conventional PCA. Clustering using K-means and PAM in the new basis set seems to produce promising results. Understanding the new basis set in the scientific context of the problem is challenging, and we are currently working to further examine and interpret the results.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2025

Landscaper v1

Understanding the inner workings of machine learning models through their loss landscapes offers crucial insights into model properties, optimization dynamics, and generalizability. However, accessing these insights has traditionally required specialized mathematical expertise, limiting broader adoption. Landscaper is an open-source Python package designed to bridge this gap. Landscaper seamlessly integrates a suite of multi-dimensional loss landscape analyses with cutting-edge topological data analysis (TDA) methods. This powerful combination makes both fundamental loss landscape analysis and advanced TDA techniques accessible to the broader scientific ML community, without requiring deep pre-existing mathematical knowledge. Landscaper offers three key functionalities: * Construction: Builds detailed loss landscape representations through versatile low and high-dimensional sampling techniques. * Quantification: Applies advanced metrics, including a novel topological data analysis (TDA) based smoothness metric, enabling new perspectives on model behavior. * Visualization: Offers intuitive tools to visualize and interpret loss landscapes, providing actionable insights beyond traditional performance metrics.

Weber, Gunther [Lawrence Berkeley National Laborat↗
NASA NTRSDate not supplied

Highlights of NASA’s Orbital Debris Program Office In Situ and Laboratory Measurements

NASA’s Orbital Debris Program Office (ODPO) maintains various returned spacecraft materials, capabilities, and facilities used for in situ and laboratory measurements that directly support orbital debris environmental models. In situ measurements include the analysis of exposed and returned hardware surfaces. These surfaces serve as passive sensors for the small-sized micrometeoroid and orbital debris (MMOD) flux below the sensitivity of ground-based radar and optical sensors. Various instruments and techniques are used to determine the size and depth of selected impact features, and – if feasible – the composition of the projectile material. Analysis of the impactor residues enables the differentiation of MM and OD for debris below 1 mm to support modeling the OD environment. In addition, projectiles identified as OD can be further differentiated in low-, medium-, and high- density impactors based on chemical analyses. In addition to in situ measurements, the ODPO has also worked in collaboration with the U.S. Space Force Space Systems Command (formerly the U.S. Air Force Space and Missile Systems Center), the Aerospace Corporation, and the University of Florida on a laboratory-based hypervelocity impact test, DebriSat, conducted at the Air Force Arnold Engineering Development Complex in 2014. The resulting data from this impact test series are being analyzed to assess the fragments’ sizes/masses, materials/densities, shapes, and other parameters of interest. The DebriSat project provides the data needed to update NASA’s breakup models and size estimation models using the simulated orbital breakup of a modern, low Earth orbit spacecraft. Ultimately, over 200,000 fragments from this impact test will be stored at NASA Johnson Space Center (JSC) and further analyzed by the ODPO. This project will also use machine learning techniques to infer physical parameters of fragments embedded in the soft-catch foam used in the impact experiment. Applied to X-ray imagery of the foam panels, these techniques promise to minimize human-in-the-loop processes for fragment extraction and physical characterization. A brief overview of this project and data collected will be presented. Lastly, the ODPO provides various capabilities hosted at NASA JSC for optical inspections and measurements using a variety of techniques and scientific instrumentation to support both in situ and laboratory measurements. The ODPO’s Optical Measurement Center (OMC) is an advanced facility for photometric and spectroscopic laboratory measurements of targets, including fragments from the DebriSat project. The OMC simulates telescopic observations by using space-like illumination conditions and source-target-sensor orientation techniques. Additionally, the OMC is uniquely equipped to acquire pseudo-bidirectional reflectance distribution data for broadband photometric measurements, thus removing aspect angle dependencies that can affect target size estimates using the optical size estimation model. Narrow-band surface material characterization using spectroscopic instrumentation gives insight into how the albedo parameter – also important in the optical size estimation model – may vary depending on the state of the material. The OMC also performs simulations of photometric measurements using optical ray-tracing software to model the OMC optical throughput. In addition to the OMC, the ODPO houses a start-of-the-art Fragment Analysis Facility that uses multiple microscopic inspection instruments to support in situ measurements and material characterization. An overview of both facilities will be highlighted in this paper.

Orbital Debris↗
NASA NTRSDate not supplied

Highlights of NASA’s Orbital Debris Program Office In Situ and Laboratory Measurements

NASA’s Orbital Debris Program Office (ODPO) maintains various returned spacecraft materials, capabilities, and facilities used for in situ and laboratory measurements that directly support orbital debris environmental models. In situ measurements include the analysis of exposed and returned hardware surfaces. These surfaces serve as passive sensors for the small-sized micrometeoroid and orbital debris (MMOD) flux below the sensitivity of ground-based radar and optical sensors. Various instruments and techniques are used to determine the size and depth of selected impact features, and – if feasible – the composition of the projectile material. Analysis of the impactor residues enables the differentiation of MM and OD for debris below 1 mm to support modeling the OD environment. In addition, projectiles identified as OD can be further differentiated in low-, medium-, and high- density impactors based on chemical analyses. In addition to in situ measurements, the ODPO has also worked in collaboration with the U.S. Space Force Space Systems Command (formerly the U.S. Air Force Space and Missile Systems Center), the Aerospace Corporation, and the University of Florida on a laboratory-based hypervelocity impact test, DebriSat, conducted at the Air Force Arnold Engineering Development Complex in 2014. The resulting data from this impact test series are being analyzed to assess the fragments’ sizes/masses, materials/densities, shapes, and other parameters of interest. The DebriSat project provides the data needed to update NASA’s breakup models and size estimation models using the simulated orbital breakup of a modern, low Earth orbit spacecraft. Ultimately, over 200,000 fragments from this impact test will be stored at NASA Johnson Space Center (JSC) and further analyzed by the ODPO. This project will also use machine learning techniques to infer physical parameters of fragments embedded in the soft-catch foam used in the impact experiment. Applied to X-ray imagery of the foam panels, these techniques promise to minimize human-in-the-loop processes for fragment extraction and physical characterization. A brief overview of this project and data collected will be presented. Lastly, the ODPO provides various capabilities hosted at NASA JSC for optical inspections and measurements using a variety of techniques and scientific instrumentation to support both in situ and laboratory measurements. The ODPO’s Optical Measurement Center (OMC) is an advanced facility for photometric and spectroscopic laboratory measurements of targets, including fragments from the DebriSat project. The OMC simulates telescopic observations by using space-like illumination conditions and source-target-sensor orientation techniques. Additionally, the OMC is uniquely equipped to acquire pseudo-bidirectional reflectance distribution data for broadband photometric measurements, thus removing aspect angle dependencies that can affect target size estimates using the optical size estimation model. Narrow-band surface material characterization using spectroscopic instrumentation gives insight into how the albedo parameter – also important in the optical size estimation model – may vary depending on the state of the material. The OMC also performs simulations of photometric measurements using optical ray-tracing software to model the OMC optical throughput. In addition to the OMC, the ODPO houses a start-of-the-art Fragment Analysis Facility that uses multiple microscopic inspection instruments to support in situ measurements and material characterization. An overview of both facilities will be highlighted in this paper.

Orbital Debris↗
DOE OSTI2025

Performance Evaluation of Intelligent Solar Control Software Through Hardware-in-the-Loop (CRADA Final Report)

Recent research has highlighted the potential for solar to act as a zero-marginal-cost and zero-emission flexibility resource on the bulk power system when operated with advanced control systems. To increase the performance of these systems, leading technologies, including machine learning (ML) and hierarchical inverter set point allocation, have been developed by Latimer Controls, Inc. to estimate the headroom of large PV plants for grid operation and control; however, these technologies lack comprehensive validation under real-world application scenarios. Latimer Controls, Inc. received two voucher awards for research at a national laboratory from the Department of Energy American Made Solar Prize Round 6. The National Renewable Energy Laboratory (NREL) was selected to collaborate with Latimer staff to conduct a performance evaluation of Latimer PV control software. The NREL team will develop a hardware-in-the-loop (HIL) testbed to perform testing and validation of the Latimer PV control technology in a de-risked yet realistic testbed environment. Latimer and NREL worked together to analyze the test data, draw conclusions from the results, and disseminate the resulting scientific findings. In this CRADA work, we propose to test and validate the real-world application of the Latimer Control solution in an HIL environment. We evaluate the performance of different flexible solar technologies in responding to automatic generation control signals in a closed-loop fashion. In particular, a data-driven potential high limit (PHL) estimation is developed for large solar plants to accurately estimate their headroom so that they have fast and short-time regulation and control capability to participate in grid services and respond to grid signals in real time (e.g., AGC). This PHL estimation algorithm is embedded in a hardware power plant controller (PPC) and tested with an IEEE-39 bus system model developed in RTDS. To account for the varying cloud conditions and diverse inverter dispatches, we developed a 135-MW PV plant with detailed modeling of 27 individual PV modules and inverters using RTDS. The real-world communications used in such big plants, such as ModBus TCP/IP for inverter level and DNP3 for plant level, were developed to emulate the real-world applications in big PV plants. The ML-based PHL estimation method is tested under nine separate weather scenarios against the ‘reference-control’ solution, hereafter referred to as the baseline solution. The baseline method reserves a subset of inverters (reference group) to operate at their PHL at all times and dispatches only the remaining inverters (control group) at curtailed levels to fulfill the flexibility need. Despite being successfully piloted by NREL in California in 2017 and Chile in 2020, there exist two gaps in the state of the art to fully unlock the flexibility of PV plants: a. There is a trade-off between the PHL estimation accuracy and the flexibility range. b. There lacks granularity in the PHL estimation to capture the variation across inverters. The Latimer solution seeks to address these gaps by applying machine learning methods to improve PHL estimation accuracy while accounting for variability at every inverter. Performance metrics were taken from the 2023 Georgia Power CARES utility-scale RFP. The results demonstrate that the ML-based approach outperforms the traditional baseline method in PHL estimation accuracy for 7 of 9 scenarios. The average PHL error across the nine scenarios was 7.40% for the ML-based method, 2.06% less than the 9.46% PHL error average across scenarios that was exhibited by the baseline method. Additionally, the PHL error was below 5% for at least 95% of the testing interval for 3 of 9 tested intervals with the ML approach, whereas it did not achieve this metric for any of the baseline tests. Overall, simulation results indicate the superior performance of an ML-based approach compared to the conventional baseline reference-control approach, showcasing its potential to support grid stability and operational efficiency. This laboratory HIL testing using real PPC, representative power system simulation models in real-time with detailed PV plant and inverter models, and real-world communication protocols gives us confidence that this machine learning based PHL estimation algorithm works well in the hardware PPC and therefore de-risks future field commissioning. The end goal of this project is to advance grid technology to address the grid operation challenges brought by solar plant’s variability and uncertainties in power generation.

14 SOLAR ENERGY↗
DOE OSTI2025

Machine Learning-Driven Conservative-to-Primitive Conversion in Hybrid Piecewise Polytropic and Tabulated Equations of State

We present a novel machine learning (ML)-based method to accelerate conservative-to-primitive inversion, focusing on hybrid piecewise polytropic and tabulated equations of state. Traditional root-finding techniques are computationally expensive, particularly for large-scale relativistic hydrodynamics simulations. To address this, we employ feedforward neural networks (NNC2PS and NNC2PL), trained in PyTorch (2.0+) and optimized for GPU inference using NVIDIA TensorRT (8.4.1), achieving significant speedups with minimal accuracy loss. The NNC2PS model achieves 𝐿 1 and 𝐿 ∞ errors of 4.54 × 10 −7 and 3.44 × 10−6, respectively, while the NNC2PL model exhibits even lower error values. TensorRT optimization with mixed-precision deployment substantially accelerates performance compared to traditional root-finding methods. Specifically, the mixed-precision TensorRT engine for NNC2PS achieves inference speeds approximately 400 times faster than a traditional single-threaded CPU implementation for a dataset size of 1,000,000 points. Ideal parallelization across an entire compute node in the Delta supercomputer (dual AMD 64-core 2.45 GHz Milan processors and 8 NVIDIA A100 GPUs with 40 GB HBM2 RAM and NVLink) predicts a 25-fold speedup for TensorRT over an optimally parallelized numerical method when processing 8 million data points. Moreover, the ML method exhibits sub-linear scaling with increasing dataset sizes. We release the scientific software developed, enabling further validation and extension of our findings. By exploiting the underlying symmetries within the equation of state, these findings highlight the potential of ML, combined with GPU optimization and model quantization, to accelerate conservative-to-primitive inversion in relativistic hydrodynamics simulations.

conservative-to-primitive conversion↗
DOE OSTI2022

Near real-time streaming analysis of big fusion data

Experiments on fusion plasmas produce high-dimensional data time series with ever-increasing magnitude and velocity, but turn-around times for analysis of this data have not kept up. For example, many data analysis tasks are often performed in a manual, ad-hoc manner some time after an experiment. In this article, we introduce the Delta framework that facilitates near real-time streaming analysis of big and fast fusion data. By streaming measurement data from fusion experiments to a high-performance compute center, Delta allows computationally expensive data analysis tasks to be performed in between plasma pulses. This article describes the modular and expandable software architecture of Delta and presents performance benchmarks of individual components as well as of an example workflow. Focusing on a streaming analysis workflow where electron cyclotron emission imaging (ECEi) data is measured at KSTAR on the National Energy Research Scientific Computing Center's (NERSC's) supercomputer we routinely observe data transfer rates of about 4 Gigabit per second. In NERSC, a demanding turbulence analysis workflow effectively utilizes multiple nodes and graphical processing units and executes them in under 5 min. We further discuss how Delta uses modern database systems and container orchestration services to provide web-based real-time data visualization. For the case of ECEi data we demonstrate how data visualizations can be augmented with outputs from machine learning models. Here, by providing session leaders and physics operators, results of higher-order data analysis using live visualizations may make more informed decisions on how to configure the machine for the next shot.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2017

Deep Learning on Operational Facility Data Related to Large-Scale Distributed Area Scientific Workflows

Abstract—Distributed computing platforms provide a robust mechanism to perform large-scale computations by splitting the task and data among multiple locations, possibly located thousands of miles apart geographically. This distribution of resources can lead to benefits such as redundancy of data, engagement with scientific teams that involve a diverse group of domain experts distributed spatially around the globe. It allows scientists to share their expertise and data with everyone and achieve insights by deploying collective intelligence of entire team. However, this distributed computing comes with its associated problems such as rampant duplication of file transfers increasing congestion, long job completion times, unexpected site crashing, suboptimal data transfer rates, unpredictable reliability in a time range, overshooting congestion, suboptimal usage of storage elements. In addition, each sub-system becomes a potential failure node that can trigger system wide disruptions.

Machine Learning, Provenance, Time Series, Semanti↗
DOE OSTI2025

Strategies for community-sourced biocuration in bioinformatics: a case study on MIBiG 4.0

Biocuration is essential to transform molecular sequence data into standardized, machine-readable resources. Such curated datasets enable comparative analysis, predictive modeling, and data integration across bioinformatics platforms. While professional biocuration is resource-intensive and usually limited to institutional settings, community-driven approaches can mobilize large-scale annotation of specialized datasets and are more resilient to disruptions in scientific funding. Here, we present a model for community-powered curation applied to the Minimum Information about a Biosynthetic Gene Cluster (MIBiG) repository. Through a framework of workflows for metadata capture, annotation validation, and contributor coordination, the MIBiG 4.0 initiative recruited 267 scientists across 178 institutions from 33 countries, volunteering an estimated 4000 h of work. These efforts expanded the MIBiG repository by 22% and enhanced its usability in downstream molecular data analyses in comparative genomic analyses, natural product discovery, and machine learning applications. We provide strategies and actionable lessons for adopting this model, supporting the sustainability of curated bioinformatics resources central to nucleic acid research and related fields.

biocuration↗
DOE OSTI2015

PANTHER. Pattern ANalytics To support High-performance Exploitation and Reasoning.

Sandia has approached the analysis of big datasets with an integrated methodology that uses computer science, image processing, and human factors to exploit critical patterns and relationships in large datasets despite the variety and rapidity of information. The work is part of a three-year LDRD Grand Challenge called PANTHER (Pattern ANalytics To support High-performance Exploitation and Reasoning). To maximize data analysis capability, Sandia pursued scientific advances across three key technical domains: (1) geospatial-temporal feature extraction via image segmentation and classification; (2) geospatial-temporal analysis capabilities tailored to identify and process new signatures more efficiently; and (3) domain- relevant models of human perception and cognition informing the design of analytic systems. Our integrated results include advances in geographical information systems (GIS) in which we discover activity patterns in noisy, spatial-temporal datasets using geospatial-temporal semantic graphs. We employed computational geometry and machine learning to allow us to extract and predict spatial-temporal patterns and outliers from large aircraft and maritime trajectory datasets. We automatically extracted static and ephemeral features from real, noisy synthetic aperture radar imagery for ingestion into a geospatial-temporal semantic graph. We worked with analysts and investigated analytic workflows to (1) determine how experiential knowledge evolves and is deployed in high-demand, high-throughput visual search workflows, and (2) better understand visual search performance and attention. Through PANTHER, Sandia's fundamental rethinking of key aspects of geospatial data analysis permits the extraction of much richer information from large amounts of data. The project results enable analysts to examine mountains of historical and current data that would otherwise go untouched, while also gaining meaningful, measurable, and defensible insights into overlooked relationships and patterns. The capability is directly relevant to the nation's nonproliferation remote-sensing activities and has broad national security applications for military and intelligence- gathering organizations.

58 GEOSCIENCES↗
NASA NTRS2021

Report on Workshop on Artificial Intelligence in Strategic Planning and Science Prioritization

This report details the observations from a two-day virtual workshop, held May 12-13, 2020, focused on whether, and how, artificial intelligence (AI) could assist humans in strategic planning, specifically in science and technology prioritization. The participants identified several “key challenges” that AI might tackle in this area. To further understand the value of these key challenges the workshop then developed related test cases that would demonstrate specifically how AI/machine learning (ML) could provide assistance to humans. Approximately 40 subject matter experts (SMEs), with backgrounds in AI, strategic planning for science, and scientific data, were gathered for the conference. This report collates the details of the output of the workshop. The “best” test cases include (in no particular order):Use of AI to assist in selecting Decadal Survey priorities. * Use of AI to identify new, or previously unidentified, science topics for prioritization. * Using AI to better label and increase discoverability of scientific literature and proposals. * Use of AI to enhance current observation capabilities for scientific missions. * Using AI to mitigate biases in selection of proposal reviewers and membership of advisory committees. Examination of these test cases indicates that Natural Language Processing (NLP) is a common capability found in most of the ”best” (top-rated) test cases and is a valuable, multi-purpose tool which enables ML in this area.

strategic planning↗
DOE OSTI2023

Convolutional Neural Network–Aided Temperature Field Reconstruction: An Innovative Method for Advanced Reactor Monitoring

In this study, the capabilities of a physics-informed convolutional neural network (CNN) for reconstructing the temperature field from a limited set of measurements taken at the boundaries of internal flows are demonstrated. Such an approach enables the development of less invasive monitoring methods for real-time plant diagnostics. As a test case, a Molten Salt Fast Reactor (MSFR) design was selected. This circulating fuel reactor has received interest from both scientific and industrial communities due to its intrinsic safety and sustainability. Molten salt flows in such reactors, however, can present highly localized temperature peaks that can induce significant thermal stresses onto the vessel walls. At these local maxima, the salt temperature may exceed a thousand kelvins, which makes a direct measurement challenging or even unfeasible. The proposed CNN algorithm allows one to detect indirectly such discontinuities through an accurate, albeit indirect, temperature measurement method during reactor operation. The datasets employed to train and test the machine learning models in the present work were generated with Nek5000, a computational fluid dynamics (CFD) code developed at Argonne National Laboratory. The CNN algorithm is trained with CFD results that span a set of MSFR operational power and flow ranges. Here, to demonstrate the efficacy of the algorithm, predictions are made for test cases contained within the training range but for which the CFD data were not used when training. Results demonstrate that the proposed technique properly characterizes temperature peaks and distributions within the domain for a broad range of scenarios.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
DOE OSTI2019

Towards Native Execution of Deep Learning on a Leadership-Class HPC System

Large parallel machines generally offer the best parallel performance with "native execution" that is achieved using codes developed with the optimized compilers, communication libraries, and runtimes offered on the machines. In this paper, we report and analyze performance results from native execution of deep learning on a leadership-class high-performance computing (HPC) system. Using our new code called DeepEx, we present a study of the parallel speed up and convergence rates of learning achieved with native parallel execution. In the trade-off between computational parallelism and synchronized convergence, we first focus on maximizing parallelism while still obtaining convergence. Scaling results are reported from execution on up to 15,000 GPUs using two scientific data sets from atom microscopy and protein folding applications, and also using the popular ImageNet data set. In terms of the traditional measure of parallel speed up, excellent scaling is observed up to 12,000 GPUs. Additionally, accounting for convergence rates of deep learning accuracy or error, a deep learning-specific metric called "learning speed up" is also tracked. The performance results indicate the need to evaluate parallel deep learning execution in terms of learning speed up, and point to additional directions for improved exploitation of high-end HPC systems.

Yoginath, Srikanth↗
DOE OSTI2020

Neural networks for modeling and control of particle accelerators

Charged particle accelerators support a wide variety of scientific, industrial, and medical applications. They range in scale and complexity from systems with just a few components for beam acceleration and manipulation, to large scientific user facilities that span many kilometers and have hundreds-to-thousands of individually-controllable components. Specific operational requirements must be met by adjusting the many controllable variables of the accelerator. Meeting these requirements can be challenging, both in terms of the ability to achieve specific beam quality metrics in a reliable fashion and in terms of the time needed to set up and maintain the optimal operating conditions. One avenue toward addressing this challenge is to incorporate techniques from the fields of machine learning (ML) and artificial intelligence (AI) into the way particle accelerators are modeled and controlled. While many promising approaches within AI/ML could be used for particle accelerators, this dissertation focuses on approaches based on neural networks. Neural networks are particularly well-suited to modeling, control, and diagnostic analysis of nonlinear systems, as well as systems with large parameter spaces. They are also very appealing for their ability to process high-dimensional data types, such as images and time series (both of which are ubiquitous in particle accelerators). In this work, key studies that demonstrated the potential utility of modern neural network-based approaches to modeling and control of particle accelerators are presented. The context for this work is important: at the start of this work in 2012, there was little interest in AI/ML in the particle accelerator community, and many of the advances in neural networks and deep learning that enabled its present success had not yet been made at that time. As such, this work was both an exploration of possible application areas and a generator of proof-of-concept demonstrations in these areas.

43 PARTICLE ACCELERATORS↗
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 OSTI2023

FRIB Transition to User Operations, Power Ramp Up, and Upgrade Perspectives

After project completion on scope, on cost, and ahead of schedule, the Facility for Rare Isotope Beams began operations for scientific users in May of 2022. During the first 12 months of user operations, the FRIB accelerator complex delivered 5250 beam hours, including 1528 hours to nine science experiments conducted with primary beams of 36Ar, 48Ca, 70Zn, 82Se, 124Xe, and 198Pt at beam energies >200 MeV/u; 2724 hours for beam developments, studies, and tuning; and 998 hours to industrial users and non-scientific programs using the FRIB Single Event Effect (FSEE) beam line. The ramp-up to a beam power of 400 kW is planned over a six-year period; 1 kW was delivered for initial user runs from in 2022, and 5 kW was delivered as of February 2023. Upgrade plans include doubling the primary-beam energy to 400 MeV/nucleon for enhanced discovery potential (¿FRIB 400¿). This talk reports on FRIB status and progress since SRF2021, emphasizing lessons learned during the transition from beam commissioning to machine operations, challenges and resolutions for the power ramp-up, progress with accelerator improvements, and R&D for the energy upgrade.

43 PARTICLE ACCELERATORS↗
NASA NTRSDate not supplied

Space Flown Rodent Liver RNA Sequencing Data for Machine Learning in Space Biology Research

High-throughput nucleic acid sequencing (DNA-seq, RNA-seq) has become widespread in biomedical research due to the growing availability and affordability of these assays. Data analysis has been accelerated in recent years by the adoption of artificial intelligence (AI) and machine learning (ML) techniques by biomedical researchers. In space biology research, RNAseq datasets from space-flown experimental samples are critical for characterizing the gene expression aberrations associated with exposure to spaceflight stressors. However, space biological experiments tend to be very low sample size, so identifying proper AI/ML algorithms for sequencing data analysis is an ongoing challenge since these algorithms typically require large sample size. The NASA Science Mission Directorate (SMD) has started the “Benchmark Initiative for AI/ML”, focused on creating datasets meant for three main applications: 1) scientific benchmarking, which finds the best algorithm for a specific problem; 2) application benchmarking, which measures algorithm performance against a set of parameters; and 3) system benchmarking, which evaluates performance of hardware and software architecture. These scientific benchmarks consist of an AI-ready dataset and a reference implementation on a specific scientific question. In this work, we focused on generating standardized datasets to allow the scientific community to benchmark AI/ML algorithms in the domain of space biology. We present here a standardized, AI-ready, publicly available benchmark dataset for space biology RNA-seq data as a collaboration between the NASA AI4LS (Artificial Intelligence for Life Sciences) working group. and NASA’s SMD. This dataset consists of space-flown and ground control mouse liver found in the NASA GeneLab omics database. However, to amplify the small sample number (n=112 samples) for ML purposes, we employ Gaussian noise and a generative adversarial network to extend this dataset to 6,000 synthetic samples, matching the original gene expression characteristics.

James Casaletto↗
DOE OSTI2021

SMC 2021 : Analyzing Resource Utilization and User Behavior on Titan Supercomputer

Resource utilization statistics of submitted jobs on a supercomputer can help us understand how users from various scientific domains use HPC platforms and better design a job scheduler. We explore to generate insight regarding workload distribution and usage pattern domains from job scheduler trace, GPU failure information, and project-specific information collected from Titan supercomputer. Furthermore, we want to know how the scheduler performance varies over time and how the users' scheduling behavior changes following a system failure. These observations have the potential to provide valuable insight, which is helpful to prepare for system failures. These practices will help us develop and apply novel machine learning algorithms in understanding system behavior, requirement, and better scheduling of HPC systems. There are two datasets, RUR and GPU. RUR: This dataset is the job scheduler traces collected from the Titan supercomputerfrom 01/01/2015 to 07/31/2019 (2015.csv - 2019.csv). These were collected usingResource Utilization Report (RUR), a Cray-developed resource-usage data collectionand reporting system. It contains the usage information of its critical resources (CPU,Memory, GPU, and I/O) of each running job on Titan during that period [2]. ProjectAreas: Every job is associated with a project ID. TheProjectAreas.csvdatasetprovides a mapping of the project ID to its domain science. GPU: There have been some hardware-related issues in the GPUs in Titan that caused some GPUs to fail, sometimes irrecoverably during some job runs. This dataset provides information regarding these failures during the execution of the submitted jobs. GPUs on Titan are uniquely identified by a serial number (SN), and they are installed in a location. A GPU can be installed in a location, then removed from that location following a failure, and then re-installed in a different location after fixing the problem. If the failure can't be recovered, the GPU might be removed entirely from Titan. There are two prominent types of failures that resulted in the removal of GPUs from Titan: Double Bit Error (DBE) and Out of the Bus (OTB). The dataset (gc_full.csv) has the following fields: 1. SN : Serial number of a GPU 2. location : The location where it is installed 3. insert : The time when it was inserted into that location 4. remove : The time when it was removed from that location 5. duration : Amount of time the GPU spent in this location 6. out : If the device was taken out entirely w/o a re-installment into a new location. 7. event : If the GPU was taken out entirely, the reason for its removal. To learn more about this dataset, please refer to the git repositoryhttps://github.com/olcf/TitanGPULifeand the related publication [1]. References [1] George Ostrouchov, Don Maxwell, Rizwan A Ashraf, Christian Engelmann, MallikarjunShankar, and James H Rogers. Gpu lifetimes on titan supercomputer: Survival analysisand reliability. InSC20: International Conference for High Performance Computing,Networking, Storage and Analysis, pages 1-14. IEEE, 2020. [2] Feiyi Wang, Sarp Oral, Satyabrata Sen, and Neena Imam. Learning from five-yearresource-utilization data of titan system. In2019 IEEE International Conference onCluster Computing (CLUSTER), pages 1-6. IEEE, 2019.

42 ENGINEERING↗