Search NASA⌕ Search

SEARCH · Search NASA

Results for “statistical 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 109 records · Page 6

Benchmarking FFTF LOFWOS Test# 13 using SAM code: Baseline model development and uncertainty quantification

The development and deployment of advanced reactors, such as the sodium-cooled fast reactor (SFR), relies on sophisticated modeling tools to ensure the safety of the design under various transients. The predictive capability of these advanced modeling tools requires validation to garner trust in supporting the licensing of the advanced reactors. For this reason, the International Atomic Energy Agency (IAEA) initiated a coordinated research project (CRP) in 2018 for the analysis of the Fast Flux Test Facility (FFTF) Loss of Flow Without Scram (LOFWOS) Test #13.In this study, we present and discuss the benchmarking efforts of the modern system code SAM on the FFTF LOFWOS Test #13. Further, the SAM baseline model was developed according to the benchmark specification, which included a detailed core model with reactivity feedback. Generally, good agreement was observed between the baseline results and benchmark measurements; however, discrepancies persisted, particularly in predicted fuel assembly coolant outlet temperatures. Utilizing the baseline model, uncertainty quantification (UQ) and sensitivity analysis (SA) were conducted with the assistance of various statistical learning and machine learning methods, including kernel density estimation, Gaussian processes, and Sobol indices. Following the baseline model prediction and UQ and SA results, we discuss the reasons for the simulation discrepancies and propose further improvements to the model. This benchmarking effort adheres to the best-estimate plus uncertainty approach and can serve as a valuable example for supporting risk-informed licensing of advanced reactors.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Matminer: An open source toolkit for materials data mining

As materials data sets grow in size and scope, the role of data mining and statistical learning methods to analyze these materials data sets and build predictive models is becoming more important. This manuscript introduces matminer, an open-source, Python-based software platform to facilitate data-driven methods of analyzing and predicting materials properties. Matminer provides modules for retrieving large data sets from external databases such as the Materials Project, Citrination, Materials Data Facility, and Materials Platform for Data Science. It also provides implementations for an extensive library of feature extraction routines developed by the materials community, with 47 featurization classes that can generate thousands of individual descriptors and combine them into mathematical functions. Finally, matminer provides a visualization module for producing interactive, shareable plots. These functions are designed in a way that integrates closely with machine learning and data analysis packages already developed and in use by the Python data science community. We explain the structure and logic of matminer, provide a description of its various modules, and showcase several examples of how matminer can be used to collect data, reproduce data mining studies reported in the literature, and test new methodologies.

97 MATHEMATICS AND COMPUTING↗

Machine learning in materials science: From explainable predictions to autonomous design

The advent of big data and algorithmic developments in the field of machine learning (and artificial intelligence, in general) have greatly impacted the entire spectrum of physical sciences, including materials science. Materials data, measured or computed, combined with various techniques of machine learning have been employed to address a myriad of challenging problems, such as, development of efficient and predictive surrogate models for a range of materials properties, screening and down-selection of novel candidate materials for targeted applications, new methodologies to improve and further expedite molecular and atomistic simulations, with likely many more important developments to come in the foreseeable future. While the applications thus far have provided a glimpse of the true potential data-enabled routes have to offer, it has also become clear that further progress in this direction hinges on our ability to understand, explain and rationalize findings of a machine learning model in light of the domain-knowledge. This focused review provides an overview of the main areas where machine learning has been widely and successfully used in materials science. Subsequently, a brief discussion of several techniques that have been helpful in extracting physically-meaningful insights, causal relationships and design-centric knowledge from materials data is provided. Finally, we identify some of the imminent opportunities and challenges that materials community faces in this exciting and rapidly growing field.

36 MATERIALS SCIENCE↗

Gradient boosting machine for modeling the energy consumption of commercial buildings

Accurate savings estimations are important to promote energy efficiency projects and demonstrate their cost-effectiveness. The increasing presence of advanced metering infrastructure (AMI) in commercial buildings has resulted in a rising availability of high frequency interval data. These data can be used for a variety of energy efficiency applications such as demand response, fault detection and diagnosis, and heating, ventilation, and air conditioning (HVAC) optimization. This large amount of data has also opened the door to the use of advanced statistical learning models, which hold promise for providing accurate building baseline energy consumption predictions, and thus accurate saving estimations. The gradient boosting machine is a powerful machine learning algorithm that is gaining considerable traction in a wide range of data driven applications, such as ecology, computer vision, and biology. In the present work an energy consumption baseline modeling method based on a gradient boosting machine was proposed. To assess the performance of this method, a recently published testing procedure was used on a large dataset of 410 commercial buildings. The model training periods were varied and several prediction accuracy metrics were used to evaluate the model's performance. The results show that using the gradient boosting machine model improved the R-squared prediction accuracy and the CV(RMSE) in more than 80 percent of the cases, when compared to an industry best practice model that is based on piecewise linear regression, and to a random forest algorithm.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Adaptive sparse polynomial chaos expansion based on least angle regression

Polynomial chaos (PC) expansions are used in stochastic finite element analysis to represent the random model response by a set of coefficients in a suitable (so-called polynomial chaos) basis. The number of terms to be computed grows dramatically with the size of the input random vector, which makes the computational cost of classical solution schemes (may it be intrusive (i.e. of Galerkin type) or non intrusive) unaffordable when the deterministic finite element model is expensive to evaluate. To address such problems, the paper describes a non intrusive method that builds a sparse PC expansion. First, an original strategy for truncating the PC expansions, based on hyperbolic index sets, is proposed. Then an adaptive algorithm based on least angle regression (LAR) is devised for automatically detecting the significant coefficients of the PC expansion. Beside the sparsity of the basis, the experimental design used at each step of the algorithm is systematically complemented in order to avoid the overfitting phenomenon. The accuracy of the PC metamodel is checked using an estimate inspired by statistical learning theory, namely the corrected leave-one-out error. As a consequence, a rather small number of PC terms are eventually retained (sparse representation), which may be obtained at a reduced computational cost compared to the classical 'full' PC approximation. The convergence of the algorithm is shown on an analytical function. Then the method is illustrated on three stochastic finite element problems. The first model features 10 input random variables, whereas the two others involve an input random field, which is discretized into 38 and 30 - 500 random variables, respectively.

97 MATHEMATICS AND COMPUTING↗

A general CFD framework for fault-resilient simulations based on multi-resolution information fusion

We develop a general CFD framework for multi-resolution simulations to target multiscale problems but also resilience in exascale simulations, where faulty processors may lead to gappy, in space-time, simulated fields. We combine approximation theory and domain decomposition together with statistical learning techniques, e.g. coKriging, to estimate boundary conditions and minimize communications by performing independent parallel runs. To demonstrate this new simulation approach, we consider two benchmark problems. First, we solve the heat equation (a) on a small number of spatial “patches” distributed across the domain, simulated by finite differences at fine resolution and (b) on the entire domain simulated at very low resolution, thus fusing multi-resolution models to obtain the final answer. Second, we simulate the flow in a lid-driven cavity in an analogous fashion, by fusing finite difference solutions obtained with fine and low resolution assuming gappy data sets. We investigate the influence of various parameters for this framework, including the correlation kernel, the size of a buffer employed in estimating boundary conditions, the coarseness of the resolution of auxiliary data, and the communication frequency across different patches in fusing the information at different resolution levels. In addition to its robustness and resilience, the new framework can be employed to generalize previous multiscale approaches involving heterogeneous discretizations or even fundamentally different flow descriptions, e.g. in continuum-atomistic simulations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Database-wide hazard modelling of the onset of DIII-D tearing modes with field features

The rate of onset (hazard) of tearing modes is modelled probabilistically using statistical learning algorithms. Axisymmetric energy-density equilibrium fields are taken as raw high-dimensional input features which are reduced with principal component analysis. Signal processing of non-axisymmetric magnetics fluctuation array data provides the target information from which to learn. Model selection, visualization and calibration assessment procedures are detailed. Here, the analysis is deployed at large scale across the DIII-D tokamak database. Standard model selection criteria suggest that the energy-density post-processed feature is a better choice for modelling the onset rate compared to the non-processed equilibrium reconstruction solution. Two example applications of the learned rate function are demonstrated: (i) proximity-to-onset discharge monitoring and (ii) database analysis showing an (expected) observational global trend that the general hazard increases as a plasma performance metric increases. An important connection between the hazard function and its use as a conditional probability generator is reviewed in the Appendix.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

High-Throughput Equilibrium Analysis of Active Materials for Solar Thermochemical Ammonia Synthesis

Solar thermochemical ammonia (NH 3 ) synthesis (STAS) is a potential route to produce NH 3 from air, water, and concentrated sunlight. This process involves the chemical looping of an active redox pair that cycles between a metal nitride and its complementary metal oxide to yield NH 3 . To identify promising candidates for STAS cycles, we performed a high-throughput thermodynamic screening of 1,148 metal nitride/metal oxide pairs. This data-driven screening was based on Gibbs energies of crystalline metal oxides and nitrides at elevated temperatures, G(T), calculated using a recently introduced statistically learned descriptor and 0 K DFT formation energies tabulated in the Materials Project database. Using these predicted G(T) values, we assessed the viability of each of the STAS reactions - hydrolysis of the metal nitride, reduction of the metal oxide, and nitrogen fixation to reform the metal nitride - and analyzed a revised cycle that directly converts between metal oxides and nitrides, which alters the thermodynamics of the STAS cycle. For all 1148 redox pairs analyzed and each of the STAS-relevant reactions, we implemented a Gibbs energy minimization scheme to predict the equilibrium composition and yields of the STAS cycle, which reveals new active materials based on B, V, Fe, and Ce that warrant further investigation for their potential to mediate the STAS cycle. This work details a high-throughput approach to assessing the relevant temperature-dependent thermodynamics of thermochemical redox processes that leverages the wealth of publicly available temperature-independent thermodynamic data calculated using DFT. This approach is readily adaptable to discovering optimal materials for targeted thermochemical applications and enabling the predictive synthesis of new compounds using thermally controlled solid-state reactions.

14 SOLAR ENERGY↗

Inorganic Halide Double Perovskites with Optoelectronic Properties Modulated by Sublattice Mixing

All-inorganic halide double perovskites have emerged as a promising class of materials that are potentially more stable and less toxic than lead-containing hybrid organic-inorganic perovskite optoelectronic materials. In this work, 311 cesium chloride double perovskites (Cs 2 BB'Cl 6 ) were selected from a set of 903 compounds as likely being stable on the basis of a statistically learned tolerance factor (t) for perovskite stability. First-principles calculations on these 311 double perovskites were then performed to assess their stability and identify candidates with band gaps appropriate for optoelectronic applications. We predict that 261 of the 311 Cs 2 BB'Cl 6 compounds are likely synthesizable on the basis of a thermodynamic analysis of their decomposition to competing compounds (decomposition enthalpy <0.05 eV/atom). Of these 261 likely synthesizable compounds, 47 contain no toxic elements and have direct or nearly direct (within 100 meV) band gaps between 1 and 3 eV, as computed with hybrid density functional theory (HSE06). Within this set, we identify the triple-alkali perovskites Cs 2 [Alk] + [TM] 3+ Cl 6 , where Alk is a group 1 alkali cation and TM is a transition-metal cation, as a class of Cs 2 BB'Cl 6 double perovskites with remarkable optical properties, including large and tunable exciton binding energies as computed by the GW-Bethe-Salpeter equation (GW-BSE) method. We attribute the unusual electronic structure of these compounds to the mixing of the Alk-Cl and TM-Cl sublattices, leading to materials with small band gaps, large exciton binding energies, and absorption spectra that are strongly influenced by the identity of the transition metal. The role of the double-perovskite structure in enabling these unique properties is probed through an analysis of the electronic structures and chemical bonding of these compounds in comparison with other transition-metal and alkali transition-metal halides.

14 SOLAR ENERGY↗

Increasing sensitivity of dryland vegetation greenness to precipitation due to rising atmospheric CO 2

Water availability plays a critical role in shaping terrestrial ecosystems, particularly in low- and mid-latitude regions. The sensitivity of vegetation growth to precipitation strongly regulates global vegetation dynamics and their responses to drought, yet sensitivity changes in response to climate change remain poorly understood. Here we use long-term satellite observations combined with a dynamic statistical learning approach to examine changes in the sensitivity of vegetation greenness to precipitation over the past four decades. We observe a robust increase in precipitation sensitivity (0.624% yr –1 ) for drylands, and a decrease (–0.618% yr –1 ) for wet regions. Using model simulations, we show that the contrasting trends between dry and wet regions are caused by elevated atmospheric CO 2 (eCO 2 ). eCO 2 universally decreases the precipitation sensitivity by reducing leaf-level transpiration, particularly in wet regions. However, in drylands, this leaf-level transpiration reduction is overridden at the canopy scale by a large proportional increase in leaf area. The increased sensitivity for global drylands implies a potential decrease in ecosystem stability and greater impacts of droughts in these vulnerable ecosystems under continued global change.

54 ENVIRONMENTAL SCIENCES↗

Exacerbated drought impacts on global ecosystems due to structural overshoot

Vegetation dynamics are affected not only by the concurrent climate but also by memory-induced lagged responses. For example, favourable climate in the past could stimulate vegetation growth to surpass the ecosystem carrying capacity, leaving an ecosystem vulnerable to climate stresses. This phenomenon, known as structural overshoot, could potentially contribute to worldwide drought stress and forest mortality but the magnitude of the impact is poorly known due to the dynamic nature of overshoot and complex influencing timescales. Here, we use a dynamic statistical learning approach to identify and characterize ecosystem structural overshoot globally and quantify the associated drought impacts. We find that structural overshoot contributed to around 11% of drought events during 1981-2015 and is often associated with compound extreme drought and heat, causing faster vegetation declines and greater drought impacts compared to non-overshoot related droughts. The fraction of droughts related to overshoot is strongly related to mean annual temperature, with biodiversity, aridity and land cover as secondary factors. These results highlight the large role vegetation dynamics play in drought development and suggest that soil water depletion due to warming-induced future increases in vegetation could cause more frequent and stronger overshoot droughts.

54 ENVIRONMENTAL SCIENCES↗

Informatics guided discovery of surface structure-chemistry relationships in catalytic nanoparticles

A data driven discovery strategy based on statistical learning principles is used to discover new correlations between electronic structure and catalytic activity of metal surfaces. From the quantitative formulations derived from this informatics based model, a high throughput computational framework for predicting binding energy as a function of surface chemistry and adsorption configuration that bypasses the need for repeated electronic structure calculations has been developed.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Hybrid PSO-ASVR-based method for data fitting in the calibration of infrared radiometer

The present paper describes a hybrid particle swarm optimization-adaptive support vector regression (PSO-ASVR)-based method for data fitting in the calibration of infrared radiometer. The proposed hybrid PSO-ASVR-based method is based on PSO in combination with Adaptive Processing and Support Vector Regression (SVR). The optimization technique involves setting parameters in the ASVR fitting procedure, which significantly improves the fitting accuracy. However, its use in the calibration of infrared radiometer has not yet been widely explored. Bearing this in mind, the PSO-ASVR-based method, which is based on the statistical learning theory, is successfully used here to get the relationship between the radiation of a standard source and the response of an infrared radiometer. Main advantages of this method are the flexible adjustment mechanism in data processing and the optimization mechanism in a kernel parameter setting of SVR. Numerical examples and applications to the calibration of infrared radiometer are performed to verify the performance of PSO-ASVR-based method compared to conventional data fitting methods.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Representing local atomic environment using descriptors based on local correlations

Statistical learning of material properties is an emerging topic of research and has been tremendously successful in areas such as representing complex energy landscapes as well as in technologically relevant areas, like identification of better catalysts and electronic materials. However, analysis of large data sets to efficiently learn characteristic features of a complex energy landscape, for example, depends on the ability of descriptors to effectively screen different local atomic environments. Thus, discovering appropriate descriptors of bulk or defect properties and the functional dependence of such properties on these descriptors remains a difficult and tedious process. To this end, we develop here a framework to generate descriptors based on many-body correlations that can effectively capture intrinsic geometric features of the local environment of an atom. These descriptors are based on the spectrum of two-body, three-body, four-body, and higher order correlations between an atom and its neighbors and are evaluated by calculating the corresponding two-body, three-body, and four-body overlap integrals. They are invariant to global translation, global rotation, reflection, and permutations of atomic indices. By systematically testing the ability to capture the local atomic environment, it is shown that the local correlation descriptors are able to successfully reconstruct structures containing 10-25 atoms which was previously not possible.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Accelerated optimization and automated discovery with covariance matrix adaptation for experimental quantum control

Optimization of quantum systems by closed-loop adaptive pulse shaping offers a rich domain for the development and application of specialized evolutionary algorithms. Derandomized evolution strategies (DESs) are presented here as a robust class of optimizers for experimental quantum control. The combination of stochastic and quasi-local search embodied by these algorithms is especially amenable to the inherent topology of quantum control landscapes. Implementation of DES in the laboratory results in efficiency gains of up to {approx}9 times that of the standard genetic algorithm, and thus is a promising tool for optimization of unstable or fragile systems. The statistical learning upon which these algorithms are predicated also provide the means for obtaining a control problem's Hessian matrix with no additional experimental overhead. The forced optimal covariance adaptive learning (FOCAL) method is introduced to enable retrieval of the Hessian matrix, which can reveal information about the landscape's local structure and dynamic mechanism. Exploitation of such algorithms in quantum control experiments should enhance their efficiency and provide additional fundamental insights.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Data-Driven Load Modeling and Forecasting of Residential Appliances

The expansion of residential demand side management programs and increased deployment of controllable loads require accurate appliance-level load modeling and forecasting. This paper proposes a conditional hidden semi-Markov model to describe the probabilistic nature of residential appliance demand. Model parameters are estimated directly from power consumption data using scalable statistical learning methods. We also propose an algorithm for short-term load forecasting as a key application for appliance-level load models. Case studies performed using granular sub-metered power measurements from various types of appliances demonstrate the effectiveness of the proposed load model for short-term prediction.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

SU-E-J-04: A Data-Driven, Response-Based, Multi-Criteria Decision Support System for Personalized Lung Radiation Treatment Planning

Purpose: To develop a decision support tool to predict a patient's potential overall survival (OS) and radiation induced toxicity (RIT) based on clinical factors and responses during the course of radiotherapy, and suggest appropriate radiation dose adjustments to improve therapeutic effect. Methods: Important relationships between a patient's basic information and their clinical features before and during the radiation treatment are identified from historical clinical data by using statistical learning and data mining approaches. During each treatment period, a data analysis (DA) module predicts radiotherapy features such as time to local progression (TTLP), time to distant metastases (TTDM), radiation toxicity to different organs, etc., under possible future treatment plans based on patient specifics or responses. An information fusion (IF) module estimates intervals for a patient's OS and the probabilities of RIT from a treatment plan by integrating the outcomes of module DA. A decision making (DM) module calculates “satisfaction” with the predicted radiation outcome based on trade-offs between OS and RIT, and finds the best treatment plan for the next time period via multi-criteria optimization. Results: Using physical and biological data from 130 lung cancer patients as our test bed, we were able to train and implement the 3 modules of our decision support tool. Examples demonstrate how it can help predict a new patient's potential OS and RIT with different radiation dose plans along with how these combinations change with dose, thus presenting a range of satisfaction/utility for use in individualized decision support. Conclusion: Although the decision support tool is currently developed from a small patient sample size, it shows the potential for the improvement of each patient's satisfaction in personalized radiation therapy. The radiation treatment outcome prediction and decision making model needs to be evaluated with more patients and demonstrated for use in radiation treatments for other cancers. P01-CA59827;R01CA142840.

60 APPLIED LIFE SCIENCES↗

Understanding GPU Memory Corruption at Extreme Scale: The Summit Case Study

GPU memory corruption and in particular double-bit errors (DBEs) remain one of the least understood aspects of HPC system reliability. Albeit rare, their occurrences always lead to job termination and can potentially cost thousands of node-hours, either from wasted computations or as the overhead from regular checkpointing needed to minimize the losses. As supercomputers and their components simultaneously grow in scale, density, failure rates, and environmental footprint, the efficiency of HPC operations becomes both an imperative and a challenge. We examine DBEs using system telemetry data and logs collected from the Summit supercomputer, equipped with 27,648 Tesla V100 GPUs with 2nd-generation high-bandwidth memory (HBM2). Using exploratory data analysis and statistical learning, we extract several insights about memory reliability in such GPUs. We find that GPUs with prior DBE occurrences are prone to experience them again due to otherwise harmless factors, correlate this phenomenon with GPU placement, and suggest manufacturing variability as a factor. On the general population of GPUs, we link DBEs to short- and long-term high power consumption modes while finding no significant correlation with higher temperatures. We also show that the workload type can be a factor in memory’s propensity to corruption.

Oles, Vlad↗