Search NASA⌕ Search

SEARCH · Search NASA

Results for “Distributed training”

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 901 records · Page 50

Evidential Deep Learning: Enhancing Predictive Uncertainty Estimation for Earth System Science Applications

Abstract Robust quantification of predictive uncertainty is a critical addition needed for machine learning applied to weather and climate problems to improve the understanding of what is driving prediction sensitivity. Ensembles of machine learning models provide predictive uncertainty estimates in a conceptually simple way but require multiple models for training and prediction, increasing computational cost and latency. Parametric deep learning can estimate uncertainty with one model by predicting the parameters of a probability distribution but does not account for epistemic uncertainty. Evidential deep learning, a technique that extends parametric deep learning to higher-order distributions, can account for both aleatoric and epistemic uncertainties with one model. This study compares the uncertainty derived from evidential neural networks to that obtained from ensembles. Through applications of the classification of winter precipitation type and regression of surface-layer fluxes, we show evidential deep learning models attaining predictive accuracy rivaling standard methods while robustly quantifying both sources of uncertainty. We evaluate the uncertainty in terms of how well the predictions are calibrated and how well the uncertainty correlates with prediction error. Analyses of uncertainty in the context of the inputs reveal sensitivities to underlying meteorological processes, facilitating interpretation of the models. The conceptual simplicity, interpretability, and computational efficiency of evidential neural networks make them highly extensible, offering a promising approach for reliable and practical uncertainty quantification in Earth system science modeling. To encourage broader adoption of evidential deep learning, we have developed a new Python package, Machine Integration and Learning for Earth Systems (MILES) group Generalized Uncertainty for Earth System Science (GUESS) (MILES-GUESS) ( https://github.com/ai2es/miles-guess ), that enables users to train and evaluate both evidential and ensemble deep learning. Significance Statement This study demonstrates a new technique, evidential deep learning, for robust and computationally efficient uncertainty quantification in modeling the Earth system. The method integrates probabilistic principles into deep neural networks, enabling the estimation of both aleatoric uncertainty from noisy data and epistemic uncertainty from model limitations using a single model. Our analyses reveal how decomposing these uncertainties provides valuable insights into reliability, accuracy, and model shortcomings. We show that the approach can rival standard methods in classification and regression tasks within atmospheric science while offering practical advantages such as computational efficiency. With further advances, evidential networks have the potential to enhance risk assessment and decision-making across meteorology by improving uncertainty quantification, a longstanding challenge. This work establishes a strong foundation and motivation for the broader adoption of evidential learning, where properly quantifying uncertainties is critical yet lacking.

Schreck, John S.↗

Machine learning snow depth predictions at sites in Alaska, Norway, Siberia, Colorado and New Mexico

Temporally continuous snow depth estimates are vital for understanding changing snow patterns in the Arctic and impacts on permafrost. We trained random forest machine learning models to predict snow depth from temperature data recorded at or just below the ground surface. Training data was collected at the Teller 27 Watershed and Kougarok 64 Hillslope during the 2021 - 2022 water year on the Seward Peninsula, Alaska using distributed temperature profiling (DTP) systems. We then applied this model to other sites where ground surface or shallow soil temperature data was available for at least one water year (see Related Datasets). Many of these temperature measurements were collocated with snow depth observations. Ground surface temperature (i.e. snow-ground interface temperature) is easy to measure using small, cheap and easy-to-deploy temperature sensors such as iButtons and TinyTags, and such measurements have previously been used to calculate a variety of snow metrics (e.g. snow onset date). However, this is the first study to estimate snow depth directly from ground surface temperature data. The present dataset contains one *.csv file which includes machine learning snow depth predictions at sites in Alaska, Norway, Siberia, Colorado, and New Mexico and one *.kml file including the locations of sites with snow depth predictions. No training data predictions are included in the *.csv file. The Next-Generation Ecosystem Experiments: Arctic (NGEE Arctic), was a research effort to reduce uncertainty in Earth System Models by developing a predictive understanding of carbon-rich Arctic ecosystems and feedbacks to climate. NGEE Arctic was supported by the Department of Energy’s Office of Biological and Environmental Research. The NGEE Arctic project had two field research sites: 1) located within the Arctic polygonal tundra coastal region on the Barrow Environmental Observatory (BEO) and the North Slope near Utqiagvik (Barrow), Alaska and 2) multiple areas on the discontinuous permafrost region of the Seward Peninsula north of Nome, Alaska. Through observations, experiments, and synthesis with existing datasets, NGEE Arctic provided an enhanced knowledge base for multi-scale modeling and contributed to improved process representation at global pan-Arctic scales within the Department of Energy’s Earth system Model (the Energy Exascale Earth System Model, or E3SM), and specifically within the E3SM Land Model component (ELM).

54 ENVIRONMENTAL SCIENCES↗

Using support vector machines for anomalous change detonation

We cast anomalous change detection as a binary classification problem, and use a support vector machine (SVM) to build a detector that does not depend on assumptions about the underlying data distribution. To speed up the computation, our SVM is implemented, in part, on a graphical processing unit. Results on real and simulated anomalous changes are used to compare performance to algorithms which effectively assume a Gaussian distribution. In this paper, we investigate the use of support vector machines (SVMs) with radial basis kernels for finding anomalous changes. Compared to typical applications of SVMs, we are operating in a regime of very low false alarm rate. This means that even for relatively large training sets, the data are quite meager in the regime of operational interest. This drives us to use larger training sets, which in turn places more of a computational burden on the SVM. We initially considered three different approaches to to address the need to work in the very low false alarm rate regime. The first is a standard SVM which is trained at one threshold (where more reliable estimates of false alarm rates are possible) and then re-thresholded for the low false alarm rate regime. The second uses the same thresholding approach, but employs a so-called least squares SVM; here a quadratic (instead of a hinge-based) loss function is employed, and for this model, there are good theoretical arguments in favor of adjusting the threshold in a straightforward manner. The third approach employs a weighted support vector machine, where the weights for the two types of errors (false alarm and missed detection) are automatically adjusted to achieve the desired false alarm rate. We have found in previous experiments (not shown here) that the first two types can in some cases work well, while in other cases they do not. This renders both approaches unreliable for automated change detection. By contrast, the third approach reliably produces good results, but at the cost of larger computational requirements caused by the need to estimate very small false alarm rates. To address these computational requirements, we employ a recently developed in-house solver for SVMs that is significantly faster than freely available standard solvers. But these computational issues are secondary to the larger question: do kernelized solutions provide better performance, in terms of detection rates and false alarm rates, than more traditional methods for change detection that effectively assume Gaussian data distributions? To this end, we will compare ROC curves obtained from the SVM with those from chronochrome, covariance equalization, and hyperbolic anomalous change detection.

97 MATHEMATICS AND COMPUTING↗

Machine-Learning Architecture for Ultrasonic Thermometry

Temperature distribution in solids can be inverted from the speed of sound (SOS) measurements, as has been shown feasible by timing the propagation of the excitation pulse and the train of echoes in ultrasonically segmented waveguides (WGs) and metal components. However, complicated geometries and closely-space echogenic features (EFs) create complex waveforms, from which the segmental time of flights (TOFs) are impossible to estimate using traditional methods. This work describes a machine learning architecture shown to extract temperature information from complex ultrasonic waveforms without explicit measurements of segmental TOFs. We accomplish this by using an autoencoder neural network (NN) to map ultrasonic waveforms into a low-dimensional latent space. A second NN then maps the latent space into unknown temperature distribution along the WG. The proposed architecture was tested in simulations and experimentally. The autoencoder accurately reconstructs the waveforms from their latent representation, several orders of magnitude lower in dimensionality. The obtained latent space was successfully mapped into the temperature of the propagation path.

John, Mason↗

Quantifying the generalization error in deep learning in terms of data distribution and neural network smoothness

We report the accuracy of deep learning, i.e., deep neural networks, can be characterized by dividing the total error into three main types: approximation error, optimization error, and generalization error. Whereas there are some satisfactory answers to the problems of approximation and optimization, much less is known about the theory of generalization. Most existing theoretical works for generalization fail to explain the performance of neural networks in practice. To derive a meaningful bound, we study the generalization error of neural networks for classification problems in terms of data distribution and neural network smoothness. We introduce the cover complexity (CC) to measure the difficulty of learning a data set and the inverse of the modulus of continuity to quantify neural network smoothness. A quantitative bound for expected accuracy/error is derived by considering both the CC and neural network smoothness. Although most of the analysis is general and not specific to neural networks, we validate our theoretical assumptions and results numerically for neural networks by several data sets of images. The numerical results confirm that the expected error of trained networks scaled with the square root of the number of classes has a linear relationship with respect to the CC. We also observe a clear consistency between test loss and neural network smoothness during the training process. In addition, we demonstrate empirically that the neural network smoothness decreases when the network size increases whereas the smoothness is insensitive to training dataset size.

97 MATHEMATICS AND COMPUTING↗

Revealing the Milky Way’s most recent major merger with a Gaia EDR3 catalogue of machine-learned line-of-sight velocities

Machine learning can play a powerful role in inferring missing line-of-sight velocities from astrometry in surveys such as Gaia. In this paper, we apply a neural network to Gaia Early Data Release 3 (EDR3) and obtain line-of-sight velocities and associated uncertainties for ~92 million stars. The network, which takes as input a star’s parallax, angular coordinates, and proper motions, is trained and validated on ~6.4 million stars in Gaia with complete phase-space information. The network’s uncertainty on its velocity prediction is a key aspect of its design; by properly convolving these uncertainties with the inferred velocities, we obtain accurate stellar kinematic distributions. As a first science application, we use the new network-completed catalogue to identify candidate stars that belong to the Milky Way’s most recent major merger, Gaia-Sausage-Enceladus (GSE). We present the kinematic, energy, angular momentum, and spatial distributions of the ~450 000 GSE candidates in this sample, and also study the chemical abundances of those with cross matches to GALAH and APOGEE. Furthermore, the network’s predictive power will only continue to improve with future Gaia data releases as the training set of stars with complete phase-space information grows. This work provides a first demonstration of how to use machine learning to exploit high-dimensional correlations on data to infer line-of-sight velocities, and offers a template for how to train, validate, and apply such a neural network when complete observational data is not available.

79 ASTRONOMY AND ASTROPHYSICS↗

Deep Learning Parameterization of Vertical Wind Velocity Variability via Constrained Adversarial Training

Atmospheric models with typical resolution in the tenths of kilometers cannot resolve the dynamics of air parcel ascent, which varies on scales ranging from tens to hundreds of meters. Small-scale wind fluctuations are thus characterized by a subgrid distribution of vertical wind velocity W with standard deviation σ W . The parameterization of σ W is fundamental to the representation of aerosol–cloud interactions, yet it is poorly constrained. Using a novel deep learning technique, this work develops a new parameterization for σ W merging data from global storm-resolving model simulations, high-frequency retrievals of W , and climate reanalysis products. The parameterization reproduces the observed statistics of σ W and leverages learned physical relations from the model simulations to guide extrapolation beyond the observed domain. Incorporating observational data during the training phase was found to be critical for its performance. The parameterization can be applied online within large-scale atmospheric models, or offline using output from weather forecasting and reanalysis products.

54 ENVIRONMENTAL SCIENCES↗

Hybrid-RL-MPC4CLR (Hybird-Reinforcement-Learning-Model-Predictive-Control-for-Reserve-Policy-Assisted-Critical-Load-Restoration-in-Distribution-Grids)

Hybrid-RL-MPC4CLR was developed as a hybrid controller for active distribution grid critical load restoration, combining deep reinforcement learning (RL) and model predictive control (MPC) aiming at maximizing total restored load following an extreme event. The RL determines a policy for quantifying operating reserve requirements, thereby hedging against uncertainty, while the MPC models grid operations incorporating the RL policy actions (i.e., reserve requirements), renewable (wind and solar) power predictions, and load demand forecasts. The developers formulated the reserve requirement determination problem as a sequential decision-making problem based on the Markov Decision Process (MDP) and design an RL learning environment based on the OpenAI Gym framework and MPC simulation. The RL agent reward and MPC objective function aim to maximize and monotonically increase total restored load and minimize load shedding and renewable power curtailment. The software is developed using various software packages in Python. The MPC's optimal power flow (OPF) model is implemented using the Pyomo package, the RL simulation environment is implemented using the MPC simulation with various scenarios of renewable energy and load demand profiles and power outage beginning times, based on the OpenAI Gym framework. The RL agent training is performed using the RLlib Ray package. The RL algorithm is trained offline using historical forecasts of renewable generation and load demand profiles. Simulation analysis and performance tests are conducted using a modified IEEE 13-bus distribution test feeder containing wind turbine, photovoltaic, microturbine, and battery.

Eseye, Abinet Tesfaye↗

Scenario adaptive disruption prediction study for next generation burning-plasma tokamaks

Next generation High Performance (HP) tokamaks risk damage from unmitigated disruptions at high current and power. Achieving reliable disruption prediction for a device's HP operation based on its Low Performance (LP) data is a key to its success. In this letter, through explorative data analysis and dedicated numerical experiments on multiple existing tokamaks, we demonstrate how the operational regimes of tokamaks can affect the power of a trained disruption predictor. First, our results suggest data-driven disruption predictors trained on abundant LP discharges work poorly on the HP regime of the same tokamak, which is a consequence of the distinct distributions of the tightly correlated signals related to disruptions in these two regimes. Second, we find that matching operational parameters among tokamaks strongly improves cross-machine accuracy which implies our model learns from the underlying scalings of dimensionless physics parameters like q 95 , β p and confirms the importance of these parameters in disruption physics and cross machine domain matching from the data-driven perspective. Finally, our results show in the absence of HP data from the target devices, the best predictivity of the HP regime for the target machine can be achieved by combining LP data from the target with HP data from other machines. Furthermore, these results provide a possible disruption predictor development strategy for next generation tokamaks, such as ITER and SPARC, and highlight the importance of developing baseline scenario discharges of future tokamaks on existing machines to collect more relevant disruptive data.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Uncertainty Quantification of Capacitor Switching Transient Location using Machine Learning

Identification of capacitor switching transient location provides valuable insight into the state of the associated equipment. Machine learning (ML) models, and convolutional neural networks (CNNs) in particular, have demonstrated remarkable performance in signal location. However, ML models are data driven whose predictions are affected by noise in data and may also suffer from large extrapolation errors when applied to new conditions. Uncertainty quantification (UQ) is necessary to ensure model trustworthiness and avoid overconfident predictions in extrapolation. Here, in this work, we propose a novel UQ method, called PI3NN, to quantify prediction uncertainty of ML models and integrate the method with CNNs for transient source location. PI3NN calculates Prediction Intervals by training 3 Neural Networks and uses root-finding methods to determine the interval precisely. Additionally, PI3NN can identify out-of-distribution (OOD) data in a nonstationary condition to avoid overconfident prediction. Results indicate that with PI3NN, transient signals are not only correctly identified, but when said signals are subject to corruptions characteristic of an actual power monitoring system (e.g. non-ideal sensors), the model recognizes when it is uncertain about its predictions, effectively letting the user know when to accept or discard the results.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Neural networks for large eddy simulations of wall-bounded turbulence: numerical experiments and challenges

Here, we examine the application of neural network-based methods to improve the accuracy of large eddy simulations of incompressible turbulent flows. The networks are trained to learn a mapping between flow features and the subgrid scales, and applied locally and instantaneously—in the same way as traditional physics-based subgrid closures. Models that use only the local resolved strain rate are poorly correlated with the actual subgrid forces obtained from filtering direct numerical simulation data. We see that highly accurate models in a priori testing are inaccurate in forward calculations, owing to the preponderance of numerical errors in implicitly filtered large eddy simulations. A network that accounts for the discretization errors is trained and found to be unstable in a posteriori testing. We identify a number of challenges that the approach faces, including a distribution shift that affects networks that fail to account for numerical errors.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

7TMRmine: a Web server for hierarchical mining of 7TMR proteins

Background: Seven-transmembrane region-containing receptors (7TMRs) play central roles in eukaryotic signal transduction. Due to their biomedical importance, thorough mining of 7TMRs from diverse genomes has been an active target of bioinformatics and pharmacogenomics research. The need for new and accurate 7TMR/GPCR prediction tools is paramount with the accelerated rate of acquisition of diverse sequence information. Currently available and often used protein classification methods (e.g., profile hidden Markov Models) are highly accurate for identifying their membership information among already known 7TMR subfamilies. However, these alignment-based methods are less effective for identifying remote similarities, e.g., identifying proteins from highly divergent or possibly new 7TMR families. In this regard, more sensitive (e.g., alignment-free) methods are needed to complement the existing protein classification methods. A better strategy would be to combine different classifiers, from more specific to more sensitive methods, to identify a broader spectrum of 7TMR protein candidates. Description: We developed a Web server, 7TMRmine, by integrating alignment-free and alignment-based classifiers specifically trained to identify candidate 7TMR proteins as well as transmembrane (TM) prediction methods. This new tool enables researchers to easily assess the distribution of GPCR functionality in diverse genomes or individual newly-discovered proteins. 7TMRmine is easily customized and facilitates exploratory analysis of diverse genomes. Users can integrate various alignment-based, alignment-free, and TM-prediction methods in any combination and in any hierarchical order. Sixteen classifiers (including two TM-prediction methods) are available on the 7TMRmine Web server. Not only can the 7TMRmine tool be used for 7TMR mining, but also for general TM-protein analysis. Users can submit protein sequences for analysis, or explore pre-analyzed results for multiple genomes. The server currently includes prediction results and the summary statistics for 68 genomes. Conclusion: 7TMRmine facilitates the discovery of 7TMR proteins. By combining prediction results from different classifiers in a multi-level filtering process, prioritized sets of 7TMR candidates can be obtained for further investigation. 7TMRmine can be also used as a general TM-protein classifier. Comparisons of TM and 7TMR protein distributions among 68 genomes revealed interesting differences in evolution of these protein families among major eukaryotic phyla.

59 BASIC BIOLOGICAL SCIENCES↗

The Challenge and Promise of Remote Work: A Brief Study of Remote Work and Best Practices

Due to the shift in labor market from manufacturing and construction to service-based (Statistics 2016); the commoditization of high-bandwidth communication tools; advancements in management techniques; a dearth of trained knowledge workers within a geographic region; and a growing demand for flexible scheduling options, organizations are embracing and encouraging the growing distributed workforce. These workers are actual employees, not contractors; they may work independently or as part of a team; they have direct organizational supervision; and they work in a variety of virtual arrangements. This informal field study looks at several organizations that support the remote work model from the perspective of individual employees and presents several technologies, tools, and policies that can serve to maintain company culture and otherwise close the gap between onsite and remote workers.

99 GENERAL AND MISCELLANEOUS↗

AOI-2, A Novel Access Control Blockchain Paradigm for Cybersecure Sensor Infrastructure in Fossil Power Generation Systems

Fossil power generation systems are increasingly vulnerable to attack from both cybercriminals as well as internal threats. These vulnerabilities demand that emerging technologies such as blockchains be utilized to secure the data involved in the information flows within the Supervisory Control and Data Acquisition (SCADA) systems of the fossil power generation plants. The publicly accessible blockchain protocols, although secure, are visible to everyone. Even private blockchains currently are unable to support different levels of access to different participants, which is a critical requirement for the existing SCADA systems running the power plants. In light of the above, novel blockchain protocols that are specifically adapted to fossil power generation environments need to be developed in order to achieve the goal of cybersecure sensor networks. In this work, we address this question by creating a novel blockchain technology, namely smart private ledger, for cybersecure communication within the fossil power generation systems. A lab-scale sensor network consisting of strain and temperature sensors is constructed to develop the ledger. The technology has hierarchical access control which is compatible with the existing SCADA systems in fossil power plants. The sensor data is used with cryptographic digital signatures and secret sharing protocols within the nodes of the blockchain technology. The research results will lead to cybersecurity for machine-to-machine interactions, infrastructure for secure data logging for sensors, decentralized data storage, and second-layer technologies for high volume machine-to-machine interactions in the power plants. The work aims to largely address the concerns for the security of distributed sensor networks in such systems that can be compromised by insider threats and by cybercriminals. The research has led to the training of the next generation of engineers and scientists in the important areas of sensor engineering and blockchain technology.

01 COAL, LIGNITE, AND PEAT↗

Super Resolution for Renewable Energy Resource Data With Wind From Reanalysis Data (Sup3rWind) and Application to Ukraine [Slides]

In this work we present a novel deep learning-based downscaling method, using generative adversarial networks (GANs), for generating high-resolution wind resource data from ECMWF Reanalysis v5 data (ERA5). We show that by training a GAN model on ERA5, as opposed to coarsened high-resolution data, we achieve results that are competitive with conventional dynamical downscaling. This GAN-based downscaling method additionally reduces computational costs over dynamical downscaling by two orders of magnitude. All GANs are trained on data sampled from CONUS, selected to provide a diverse sampling of terrain conditions, and validated on observational data along with data held out from training. This cross-validation shows low error and high correlations with observations and excellent agreement with hold out data across physical distributions. Our approach is finally used to downscale 30km hourly ERA5 to 2-km 5-minute wind data, for January 2000 through December 2023, at multiple hub heights, over Ukraine, Moldova, and part of Romania. Comparisons against observational data from Meteorological Assimilation Data Ingest System (MADIS) and multiple wind farms show the same level of performance as for CONUS validation. This 24 year data record is the first member of the "super resolution for renewable energy resource data with wind from reanalysis data" dataset (Sup3rWind).

17 WIND ENERGY↗

Detecting Anomalous Images in Astronomical Datasets

Abstract Environmental and instrumental conditions can cause anomalies in astronomical images, which can potentially bias all kinds of measurements if not excluded. Detection of the anomalous images is usually done by human eyes, which is slow and sometimes not accurate. This is an important issue in weak lensing studies, particularly in the era of large-scale galaxy surveys, in which image qualities are crucial for the success of galaxy shape measurements. In this work we present two automatic methods for detecting anomalous images in astronomical data sets. The anomalous features can be divided into two types: one is associated with the source images, and the other appears on the background. Our first method, called the entropy method, utilizes the randomness of the orientation distribution of the source shapes and the background gradients to quantify the likelihood of an exposure being anomalous. Our second method involves training a neural network (autoencoder) to detect anomalies. We evaluate the effectiveness of the entropy method on the Canada–France–Hawaii Telescope Lensing Survey (CFHTLenS) and Dark Energy Camera Legacy Survey (DECaLS DR3) data. In CFHTLenS, with 1171 exposures, the entropy method outperforms human inspection by detecting 12 of the 13 anomalous exposures found during human inspection and uncovering 10 new ones. In DECaLS DR3, with 17112 exposures, the entropy method detects a significant number of anomalous exposures while keeping a low false-positive rate. We find that although the neural network performs relatively well in detecting source anomalies, its current performance is not as good as the entropy method.

Astronomy & Astrophysics↗

PV Inverter Experimental Dataset Version 2 with 100 Percent Power

The increase in power electronic based generation sources require accurate modeling of inverters. Accurate modeling requires experimental data over wider operation range. We used 20 kW off-the-shelf grid following PV inverter in the experiments. We used controllable AC supply and controllable DC supply to emulate AC and DC side characteristics. The experiments were performed at NREL's Energy Systems Integration Facility. The PV inverter is tested under 100%, 75%, 50%, 25% load conditions. In the first dataset, for each operating condition, controllable AC source voltage is varied from 0.88 to 1.09 per unit (p.u) with a step value of 0.025 p.u while keeping the frequency at 60 Hz. In the second dataset, under similar load conditions (100%, 75%, 50%, 25% ), the frequency of the controllable AC source voltage was varied from 59.4 Hz to 60.45 Hz with a step value of 0.2 Hz. Voltage and frequency range is chosen based on inverter protection. Voltages and currents on DC and AC side are included in the dataset.

14 SOLAR ENERGY↗

Application of multi-function display and control technology

The NASA orbiter spacecraft incorporates a complex array of systems, displays, and controls. The incorporation of discrete dedicated controls into a multifunction display and control system (MFDCS) offers the potential for savings in weight, power, panel space, and crew training time. Technology identified as applicable to a MFDCS is applied to the orbiter orbital maneuvering system (OMS) and the electrical power distribution and control system (EPDCS) to derive concepts for a MFDCS design. Several concepts of varying degrees of performance and complexity are discussed and a suggested concept for further development is presented in greater detail. Both the hardware and software aspects and the human factors considerations of the designs are included.

Spiger, R. J.↗