Search NASA⌕ Search

SEARCH · Search NASA

Results for “randomized sampling”

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 811 records · Page 45
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

The hidden structure of hydrodynamic transport in random fracture networks

We study the large-scale dynamics and prediction of hydrodynamic transport in random fracture networks. The flow and transport behaviour is characterized by first passage times and displacement statistics, which show heavy tails and anomalous dispersion with a strong dependence on the injection condition. The origin of these behaviours is investigated in terms of Lagrangian velocities sampled equidistantly along particle trajectories, unlike classical sampling strategies at a constant rate. The velocity series are analysed by their copula density, the joint distribution of the velocity unit scores, which reveals a simple, albeit hidden, correlation structure that can be described by a Gaussian copula. Based on this insight, we derive a Langevin equation for the evolution of equidistant particle speeds. In this framework, particle motion is quantified by a stochastic time-domain random walk, the joint density of particle position, and speed satisfies a Klein–Kramers equation. The upscaled theory quantifies particle motion in terms of the characteristic fracture length scale and the distribution of Eulerian flow velocities. That is, it is predictive in the sense that it does not require the a priori knowledge of transport attributes. The upscaled model captures non-Fickian transport features, and their dependence on the injection conditions in terms of the velocity point statistics and average fracture length. It shows that the first passage times and displacement moments are dominated by extremes occurring at the first step. The presented approach integrates the interaction of flow and structure into a predictive model for large-scale transport in random fracture networks.

42 ENGINEERING↗
DOE OSTI2016

Controlled release of cortisone drugs from block copolymers synthetized by ATRP

Diseases affecting posterior eye segment, like macular edema, infection and neovascularization, may cause visual impairment. Traditional treatments, such as steroidal-drugs intravitreal injections, involve chronic course of therapy usually over a period of years. Moreover, they can require frequent administrations of drug in order to have an adequately disease control. This dramatically reduce patient’s compliance. Efforts have been made to develop implantable devices that offer an alternative therapeutic approach to bypass many challenges of conventional type of therapy. Implantable drug delivery systems (DDS) have been developed to optimize therapeutic properties of drugs and ensure their slow release in the specific site. Polymeric materials can play an essential role in modulating drug delivery and their use in such field has become indispensable. During last decades, acrylic polymers have obtained growing interest. Biocompatibility and chemical properties make them extremely versatile, allowing their use in many field such as biomedical. In particular, block methacrylate copolymer with a balance of hydrophilic and hydrophobic properties can be suitable for prolonged DDS in biomedical devices. In this work, we focused on the realization of a system for controlled and long term release of betamethasone 17,21-dipropionate (BDP), a cortisone drug, from methacrylic block copolymers, to be tested in the treatment of the posterior eye’s diseases. Different series of methyl methacrylate/hydroxyethyl methacrylate (MMA/HEMA) block and random copolymers, with different monomer compositions (10–60% HEMA), were synthetized by Atom Transfer Radical Polymerization (ATRP) to find the best hydrophilic/hydrophobic ratio, able to ensure optimal kinetic release. Copolymer samples were characterized by NMR spectroscopy ({sup 1}H-NMR, {sup 13}C-NMR, CosY), SEC, TGA and DSC. Monitoring of drug release from films loaded with BDP was carried out by HPLC analysis. Evaluation of different kinetic models allowed to deduce that release of BDP is controlled over time from PMMA-b-PHEMA 53/47. In particular, PMMA-b-PHEMA 53/47 showed the best release profile to achieve the therapeutic reference dose of 3 µg/die, employed in treatment of posterior eye disease, up to four months. Accordingly, PMMA-b-PHEMA 53/47 has been tested to prepare ocular inserts. Ocular inserts with different shape and the same area of polymer films have been obtained using silicon moulds made by a 3D printer.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2016

Ion irradiation-induced crystal structure changes in inverse spinel MgIn 2 O 4

We performed 400 keV Ne and 200 keV He ion irradiations on fully inverse MgIn 2 O 4 samples at cryogenic temperature (~ 77 K), in order to examine the influence of radiation-induced cation disordering on crystal structure. In the case of MgIn 2 O 4 samples irradiated with Ne ions to a peak displacement damage dose of 4 displacements per atom (dpa), a spinel-to-rocksalt phase transformation was observed. Conversely, for MgIn 2 O 4 samples irradiated with He ions to a peak displacement damage dose of 5 dpa, the only observed structural effect involved cation rearrangements from an inverse to a “random” spinel structure.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Observed Faraday Effects in Damped Lyα Absorbers and Lyman Limit Systems: The Magnetized Environment of Galactic Building Blocks at Redshift = 2

Protogalactic environments are typically identified using quasar absorption lines and can manifest as Damped Lyman-alpha Absorbers (DLAs) and Lyman Limit Systems (LLSs). We use radio observations of Faraday effects to test whether these galactic building blocks host a magnetized medium, by combining DLA and LLS detections with 1.4 GHz polarization data from the NRAO VLA Sky Survey (NVSS). We obtain a control, a DLA, and an LLS sample consisting of 114, 19, and 27 lines of sight, respectively. Using a Bayesian framework and weakly informative priors, we are unable to detect either coherent or random magnetic fields in DLAs: the regular coherent fields must be ⩽2.8 μG, and the lack of depolarization suggests the weakly magnetized gas in DLAs is non-turbulent and quiescent. However, we find a mild suggestive indication that LLSs have coherent magnetic fields, with a 71.5% probability that LLSs have higher |RM| than a control, although this is sensitive to the redshift distribution. We also find a strong indication that LLSs host random magnetic fields, with a 95.5% probability that LLS lines of sight have lower polarized fractions than a control. The regular coherent fields within the LLSs must be ⩽2.4 μG, and the magnetized gas must be highly turbulent with a typical turbulent length scale on the order of ≈5–20 pc. Our results are consistent with the standard dynamo paradigm, whereby magnetism in protogalaxies increases in coherence over cosmic time, and with a hierarchical galaxy formation scenario, with the DLAs and LLSs exploring different stages of magnetic field evolution in galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Computer-Based Procedures Among Older Adults: Findings from the Rancor Microworld Simulator

As with many industries worldwide, nuclear energy is experiencing an aging workforce; older than other energy sources and the national average. For example, while almost one in three nuclear industry professionals are 55+ years, for oil and gas that number is one in five. The challenges brought about by an aging workforce are one of the industry’s top concerns. However, the industry lacks for empirical research that examines the effects of developmental aging in older workers, as well as the interactions between aging and new digital technologies. We present the results of an experiment that tested three different computer-based procedures in a sample of 30 older adults (55+ years). While the industry has traditionally relied on paper-based procedures, these are being modernized by digital technology. Participants were randomly assigned to one of three procedure-types that varied by the level of digitalization, based on the IEEE Standard-1786. Type 1 essentially represents a digital representation of a paper-based procedure, Type 2 adds embedded indicators, and Type 3 adds soft controls. Participants performed two different operational scenarios (startup and loss of feedwater) on a simplified nuclear power plant simulator. Results revealed a weak signal that Type 2 may produce lower workload and lower completion times in some instances. However, there were no significant effects of procedure-type across any other outcomes, including simulator log data, situation awareness, and preference ratings. We discuss our findings in terms of optimal levels of digitalization/automation for an aging nuclear workforce and suggest pathways for future directions.

46 - INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AN↗
DOE OSTI2023

Quantum Search Approaches to Sampling-Based Motion Planning

In this paper, we present a novel formulation of traditional sampling-based motion planners as database-oracle structures that can be solved via quantum search algorithms. We consider two complementary scenarios: for simpler sparse environments, we formulate the Quantum Full Path Search Algorithm (q-FPS), which creates a superposition of full random path solutions, manipulates probability amplitudes with Quantum Amplitude Amplification (QAA), and quantum measures a single obstacle free full path solution. For dense unstructured environments, we formulate the Quantum Rapidly Exploring Random Tree algorithm, q-RRT, that creates quantum superpositions of possible parent-child connections, manipulates probability amplitudes with QAA, and quantum measures a single reachable state, which is added to a tree. As performance depends on the number of oracle calls and the probability of measuring good quantum states, we quantify how these errors factor into the probabilistic completeness properties of the algorithm. We then numerically estimate the expected number of database solutions to provide an approximation of the optimal number of oracle calls in the algorithm. We compare the q-RRT algorithm with a classical implementation and verify quadratic run-time speedup in the largest connected component of a 2D dense random lattice. We conclude by evaluating a proposed approach to limit the expected number of database solutions and thus limit the optimal number of oracle calls to a given number.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Predicting oxidation damage in ultra high-temperature borides: A machine learning approach

Ultra-high temperature (UHT) borides are ceramics materials with melting points above 3000 °C for structural applications in extreme environments. However, at temperatures exceeding 1600 °C and under oxidizing conditions, the material suffers from detrimental degradation. Optimized design and performance of diboride materials under such extreme conditions requires filling the missing composition-microstructure-oxidation gap. This study proposes a computational data-driven framework to connect the processing and microstructure of Ultra-high temperature borides with the oxidation damage. Random Forest Regressor (RFR) model is adopted to forecast the oxide scale thickness developed after oxidation testing based on processing variables and microstructural features. The model trained on a dataset consisting of 107 samples of experimental data extracted from the literature aims to predict oxidation damage. With proper data manipulation and fine model tuning, the predictor could forecast the oxide scale thickness of UHT diborides with a Mean Absolute Error of 37.45 μm and an R-square of 0.83. This model could be used as a high-throughput scheme to design and test new UHT diborides materials computationally. Furthermore, a model with larger composition capabilities could also be developed in the future as more experimental data become available.

36 MATERIALS SCIENCE↗
DOE OSTI2019

AC Magnetic Susceptibility of Y{sub 3}Ba{sub 5}Cu{sub 8}O{sub 18} Substituted by Nd{sup 3+} and Ca{sup 2+} Ions

Effect of Nd{sup 3+} and Ca{sup 2+} ion substitutions in Y{sub 3}Ba{sub 5}Cu{sub 8}O{sub 18} (Y-358) was studied using X-ray diffraction (XRD), scanning electron microscopy (SEM), electrical resistivity, and ac magnetic susceptibility measurements. Superconducting samples of type Y{sub 3 − x}Nd{sub x}Ba{sub 5 − y}Ca{sub y}Cu{sub 8}O{sub 18} (with 0 ≤ x, y ≤ 0.4) were synthesized by the standard solid-state reaction technique. XRD patterns revealed that all the samples were mainly composed of Y-358 phase with orthorhombic structure. SEM micrographs showed that Nd{sup 3+} and Ca{sup 2+} ion substitutions had the ability to alter the grain morphology of Y-358 to irregular randomly oriented grains with smaller average grain sizes. The temperature-dependent electrical resistivity curves indicated that the superconducting transition temperature T{sub c} was decreased from 96.51 to 87.41 K as the substitution contents were increased from 0 to 0.4. Detailed AC magnetic susceptibility measurements were done using a quantum design physical property measurement system (PPMS) at different dc fields (B = 0.5–20 kG) and frequency of 1 kHz. A comparison between electrical resistivity and AC magnetic susceptibility data was performed. It was found that T{sub cs} obtained from AC magnetic susceptibility were much lower than those obtained from the resistivity measurements, for all studied samples. This was discussed in terms of filamentary paths that could produce sharp pronounced drops in diamagnetism. Generally, the depression in T{sub c} was attributed to the decrease in relative volume fraction of Y-358 or due to trapping of mobile holes or other mechanism connected with oxygen vacancy disorder. Werthamer–Helfand–Hohenberg (WHH) theory was employed to study the temperature dependence of the upper critical magnetic field B{sub c2}(T) for the samples. The relative importance of both orbital and paramagnetic effects in the suppression of superconductivity was described with the aid of Maki parameter (κ{sub M}). The estimated values of the upper critical magnetic field at 0 K, B{sub c2}(0), were found to decrease from 37.3 to 13.5 T as Nd{sup 3+} and Ca{sup 2+} contents were increased from 0 to 0.4. This decrement was attributed to the weak electron–phonon coupling and to the orbital as well as spin pair breaking mechanisms. Furthermore, Ginzburg–Landau coherence length ξ{sub GL} of Y{sub 3 − x}Nd{sub x}Ba{sub 5 − y}Ca{sub y}Cu{sub 8}O{sub 18} was found to increase with the increase of the Nd{sup 3+} and Ca{sup 2+} contents. Therefore, Nd{sup 3+} and Ca{sup 2+} ions play an important role in the pair-breaking mechanism and in suppressing superconductivity of the Y-358 phase.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Ensemble methods for quantification of potassium oxide in ChemCam Mars and laboratory spectra

In this paper we test new approaches for predicting the amount of element oxides in rock samples from the ChemCam instrument suite onboard the NASA Curiosity rover by focusing on K 2 O. Using the expanded dataset compiled by Gasda et al. (2021) with and without the Earth to Mars (E2M and NoE2M) transformation discussed in Clegg et al. (2017) we trained blended submodels using the “double blending” technique and compared these to ensemble methods (Random Forest, ExtraTrees, and Gradient Boosting Regression). We found that ensemble methods performed similar to blended submodels when looking at RMSE-P on the laboratory spectra and provided significant advantages when looking at spectra coming from Mars. For the full model, blended submodels achieved an RMSE-P of 0.62 and 0.60 (E2M and NoE2M respectively) while Gradient Boosting Regression resulted in a slightly improved RMSE-P of 0.59 and 0.60. More importantly, by employing a local RMSE-P estimation technique where model performance is evaluated based on nearby test samples we found that using ensemble methods can lower the quantification limit for K 2 O from the current value of ≈0.6 wt% to ≈0.08 wt% using Extra Trees and Random Forest. This would allow for a much larger range of K 2 O values to be quantified on Mars with greater certainty given that most targets seen on Mars tend to have <1 wt% K2O. Finally, we used both Mean Decrease in Impurity (MDI) and permutation importance techniques to investigate the wavelengths used by the ensemble methods and found that they correspond to known potassium emission lines. This suggests that ensemble methods can provide an easier to train and improved alternative to blended submodels for predicting potassium compositions from Laser Induced Breakdown Spectroscopy (LIBS) data.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

The hyperplane of early-type galaxies: using stellar population properties to increase the precision and accuracy of the fundamental plane as a distance indicator

ABSTRACT We use deep spectroscopy from the SAMI (Sydney-AAO Multi-object Integral) Galaxy Survey to explore the precision of the fundamental plane (FP) of early-type galaxies as a distance indicator for future single-fibre spectroscopy surveys. We study the optimal trade-off between sample size and signal-to-noise ratio (SNR), and investigate which additional observables can be used to construct hyperplanes with smaller intrinsic scatter than the FP. We add increasing levels of random noise (parametrized as effective exposure time) to the SAMI spectra to study the effect of increasing measurement uncertainties on the FP- and hyperplane-inferred distances. We find that, using direct-fit methods, the values of the FP and hyperplane best-fitting coefficients depend on the spectral SNR, and reach asymptotic values for a mean $\langle \mathrm{ SNR} \rangle =40\, \mathrm{\mathring{\rm A}}^{-1}$. As additional variables for the FP we consider three stellar-population observables: light-weighted age, stellar mass-to-light ratio, and a novel combination of Lick indices ($I_\mathrm{age}$). For an $\langle \mathrm{ SNR} \rangle =45~\mathrm{\mathring{\rm A}}^{-1}$ (equivalent to 1-h exposure on a 4-m telescope), all three hyperplanes outperform the FP as distance indicators. Being an empirical spectral index, $I_\mathrm{age}$ avoids the model-dependent uncertainties and bias underlying age and mass-to-light ratio measurements, yet yields a 10 per cent reduction of the median distance uncertainty compared to the FP. We also find that, as a by-product, the $I_\mathrm{age}$ hyperplane removes most of the reported environment bias of the FP. After accounting for the different SNR, these conclusions also apply to a 50 times larger sample from SDSS-III (Sloan Digital Sky Survey). However, in this case, only $\mathrm{ age}$ removes the environment bias.

D’Eugenio, Francesco (ORCID:0000000323888172)↗
DOE OSTI2019

Disorder-to-Order Transition Mediated by Size Refocusing: A Route toward Monodisperse Intermetallic Nanoparticles

Intermetallic nanoparticles are remarkable because of their often enhanced catalytic, magnetic, and optical properties, which arise from their ordered crystal structures and high structural stability. Typical syntheses of intermetallic nanoparticles include thermal annealing of the disordered counterpart in atmosphere (or vacuum) or colloidal syntheses, where the phase transformation is achieved in solution. Although both methods can produce intermetallic nanoparticles, there is difficulty in achieving monodisperse nanoparticles, which is critical to exploiting their properties for various applications. In this work, we demonstrate that overgrowth on random alloy AuCu nanoparticles mediated by size refocusing yields monodisperse intermetallic AuCu nanoparticles. Size refocusing has been used in syntheses of semiconductor and upconverting nanocrystals to achieve monodisperse samples, but now we demonstrate size refocusing as a mechanism to achieve the disorder-to-order phase transformation in multimetallic nanoparticles. The phase transformation was monitored by time evolution experiments, where analysis of reaction aliquots with transmission electron microscopy and powder X-ray diffraction revealed the generation and dissolution of small nanoparticles coupled with an increase in the average size of the nanoparticles and conversion to the ordered phase. This demonstration advances the understanding of intermetallic nanoparticle formation in colloidal syntheses, which can expedite the development of electrocatalysts and magnetic storage materials.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2020

Machine learning prediction of incidence of Alzheimer’s disease using large-scale administrative health data

Nationwide population-based cohort provides a new opportunity to build an automated risk prediction model based on individuals’ history of health and healthcare beyond existing risk prediction models. We tested the possibility of machine learning models to predict future incidence of Alzheimer’s disease (AD) using large-scale administrative health data. From the Korean National Health Insurance Service database between 2002 and 2010, we obtained de-identified health data in elders above 65 years (N = 40,736) containing 4,894 unique clinical features including ICD-10 codes, medication codes, laboratory values, history of personal and family illness and socio-demographics. To define incident AD we considered two operational definitions: “definite AD” with diagnostic codes and dementia medication (n = 614) and “probable AD” with only diagnosis (n = 2026). We trained and validated random forest, support vector machine and logistic regression to predict incident AD in 1, 2, 3, and 4 subsequent years. For predicting future incidence of AD in balanced samples (bootstrapping), the machine learning models showed reasonable performance in 1-year prediction with AUC of 0.775 and 0.759, based on “definite AD” and “probable AD” outcomes, respectively; in 2-year, 0.730 and 0.693; in 3-year, 0.677 and 0.644; in 4-year, 0.725 and 0.683. The results were similar when the entire (unbalanced) samples were used. Important clinical features selected in logistic regression included hemoglobin level, age and urine protein level. This study may shed a light on the utility of the data-driven machine learning model based on large-scale administrative health data in AD risk prediction, which may enable better selection of individuals at risk for AD in clinical trials or early detection in clinical settings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Window convolution of the galaxy clustering bispectrum

In galaxy survey analysis, the observed clustering statistics do not directly match theoretical predictions but rather have been processed by a window function that arises from the survey geometry including the sky footprint, redshift-dependent background number density and systematic weights. While window convolution of the power spectrum is well studied, for the bispectrum with a larger number of degrees of freedom, it poses a significant numerical and computational challenge. In this work, we consider the effect of the survey window in the tripolar spherical harmonic decomposition of the bispectrum and lay down a formal procedure for their convolution via a series expansion of configuration-space three-point correlation functions, which was first proposed by Sugiyama et al. (2019). We then provide a linear algebra formulation of the full window convolution, where an unwindowed bispectrum model vector can be directly premultiplied by a window matrix specific to each survey geometry. To validate the pipeline, we focus on the Dark Energy Spectroscopic Instrument (DESI) Data Release 1 (DR1) luminous red galaxy (LRG) sample in the South Galactic Cap (SGC) in the redshift bin 0.4 ≤ z ≤ 0.6. We first perform convergence checks on the measurement of the window function from discrete random catalogues, and then investigate the convergence of the window convolution series expansion truncated at a finite of number of terms as well as the performance of the window matrix. This work highlights the differences in window convolution between the power spectrum and bispectrum, and provides a streamlined pipeline for the latter for current surveys such as DESI and the Euclid mission.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Neutral but Impactful: Gallium Cluster-Induced Nanopores from Beam-Blanked Gallium Ion Sources

Neutral atoms originating from liquid metal ion sources are an often-overlooked source of contamination and damage in focused ion beam microscopy. Beyond ions and single atoms, these sources also generate atom clusters. While most studies have investigated charged clusters, here we demonstrate that neutral clusters are also formed. These neutral clusters bypass the electrostatic beam blanking system, allowing them to impinge on samples even when the ion beam is blanked. We investigate this phenomenon using thin (≤20 nm) freestanding membranes of hexagonal boron nitride, silicon, and silicon nitride as targets. Randomly dispersed nanopores that form upon neutral cluster exposure are revealed. The average nanopore diameter is ∼2 nm with a narrow size distribution, suggesting that the atom clusters have a preferred size. Various electron microscopy techniques are used to characterize the nanopores, including high-resolution transmission electron microscopy, multislice ptychography, and electron energy-loss spectroscopy. Finally, we show how electron irradiation in the transmission electron microscope can be used to both remove any amorphous material that may clog the pores and to controllably grow the pores to specific sizes. Tunable nanopores such as these are interesting for nanofluidic applications requiring size-selective membranes.

Byrne, Dana O↗
DOE OSTI2018

Photometric classification and redshift estimation of LSST Supernovae

Supernova (SN) classification and redshift estimation using photometric data only have become very important for the Large Synoptic Survey Telescope (LSST), given the large number of SNe that LSST will observe and the impossibility of spectroscopically following up all the SNe. We investigate the performance of an SN classifier that uses SN colours to classify LSST SNe with the Random Forest classification algorithm. Our classifier results in an area-under-the-curve of 0.98 which represents excellent classification. We are able to obtain a photometric SN sample containing 99 percent SNe Ia by choosing a probability threshold. We estimate the photometric redshifts (photo-z) of SNe in our sample by fitting the SN light curves using the SALT2 model with nested sampling. We obtain a mean bias ($\langle$z phot - z spec $\rangle$) of 0.012 with σ($\frac{zphot-zspec}{1+zspec}$)=0.0294 σ(zphot-zspec1+zspec)=0.0294 without using a host-galaxy photo-z prior, and a mean bias ($\langle$z phot - z spec $\rangle$) of 0.0017 with σ($\frac{zphot-zspec}{1+zspec}$)=0.0116 using a host-galaxy photo-z prior. Assuming a flat ΛCDM model with Ω m = 0.3, we obtain Ω m of 0.305 ± 0.008 (statistical errors only), using the simulated LSST sample of photometric SNe Ia (with intrinsic scatter σ int = 0.11) derived using our methodology without using host-galaxy photo-z prior. Our method will help boost the power of SNe from the LSST as cosmological probes.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Evaluating the impact of anatomical and physiological variability on human equivalent doses using PBPK models

Abstract Addressing human anatomical and physiological variability is a crucial component of human health risk assessment of chemicals. Experts have recommended probabilistic chemical risk assessment paradigms in which distributional adjustment factors are used to account for various sources of uncertainty and variability, including variability in the pharmacokinetic behavior of a given substance in different humans. In practice, convenient assumptions about the distribution forms of adjustment factors and human equivalent doses (HEDs) are often used. Parameters such as tissue volumes and blood flows are likewise often assumed to be lognormally or normally distributed without evaluating empirical data for consistency with these forms. In this work, we performed dosimetric extrapolations using physiologically based pharmacokinetic (PBPK) models for dichloromethane (DCM) and chloroform that incorporate uncertainty and variability to determine if the HEDs associated with such extrapolations are approximately lognormal and how they depend on the underlying distribution shapes chosen to represent model parameters. We accounted for uncertainty and variability in PBPK model parameters by randomly drawing their values from a variety of distribution types. We then performed reverse dosimetry to calculate HEDs based on animal points of departure for each set of sampled parameters. Corresponding samples of HEDs were tested to determine the impact of input parameter distributions on their central tendencies, extreme percentiles, and degree of conformance to lognormality. This work demonstrates that the measurable attributes of human variability should be considered more carefully and that generalized assumptions about parameter distribution shapes may lead to inaccurate estimates of extreme percentiles of HEDs.

Toxicology↗
DOE OSTI2019

Data-Consistent Solutions to Stochastic Inverse Problems Using a Probabilistic Multi-Fidelity Method Based on Conditional Densities

In this work, we build upon a recently developed approach for solving stochastic inverse problems based on a combination of measure-theoretic principles and Bayes' rule. We propose a multi-fidelity method to reduce the computational burden of performing uncertainty quantification using high-fidelity models. This approach is based on a Monte Carlo framework for uncertainty quantification that combines information from solvers of various fidelities to obtain statistics on the quantities of interest of the problem. In particular, our goal is to generate samples from a high-fidelity push-forward density at a fraction of the costs of standard Monte Carlo methods, while maintaining flexibility in the number of random model input parameters. Key to this methodology is the construction of a regression model to represent the stochastic mapping between the low- and high-fidelity models, such that most of the computations can be leveraged to the low-fidelity model. To that end, we employ Gaussian process regression and present extensions to multi-level-type hierarchies as well as to the case of multiple quantities of interest. Finally, we demonstrate the feasibility of the framework in several numerical examples.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Extreme-Scale Bayesian Inference for Uncertainty Quantification of Complex Simulations

Uncertainty quantification (UQ)—that is, quantifying uncertainties in complex mathematical models and their large-scale computational implementations—is widely viewed as one of the outstanding challenges facing the field of CS&E over the coming decade. The EUREKA project set to address the most difficult class of UQ problems: those for which both the underlying PDE model as well as the uncertain parameters are of extreme scale. In the project we worked on these extreme-scale challenges in the following four areas: 1. Scalable parallel algorithms for sampling and characterizing the posterior distribution that exploit the structure of the underlying PDEs and parameter-to-observable map. These include structure-exploiting versions of the randomized maximum likelihood method, which aims to overcome the intractability of employing conventional MCMC methods for solving extreme-scale Bayesian inversion problems by appealing to and adapting ideas from large-scale PDE-constrained optimization, which have been very successful at exploring high-dimensional spaces. 2. Scalable parallel algorithms for construction of prior and likelihood functions based on learning methods and non-parametric density estimation. Constructing problem-specific priors remains a critical challenge in Bayesian inference, and more so in high dimensions. Another challenge is construction of likelihood functions that capture unmodeled couplings between observations and parameters. We will create parallel algorithms for non-parametric density estimation using high dimensional N-body methods and combine them with supervised learning techniques for the construction of priors and likelihood functions. 3. Bayesian inadequacy models, which augment physics models with stochastic models that represent their imperfections. The success of the Bayesian inference framework depends on the ability to represent the uncertainty due to imperfections of the mathematical model of the phenomena of interest. This is a central challenge in UQ, especially for large-scale models. We propose to develop the mathematical tools to address these challenges in the context of extreme-scale problems. 4. Parallel scalable algorithms for Bayesian optimal experimental design (OED). Bayesian inversion yields quantified uncertainties in the model parameters, which can be propagated forward through the model to yield uncertainty in outputs of interest. This opens the way for designing new experiments to reduce the uncertainties in the model parameters and model predictions. Such experimental design problems have been intractable for large-scale problems using conventional methods; we will create OED algorithms that exploit the structure of the PDE model and the parameter-to-output map to overcome these challenges. Parallel algorithms for these four problems were created, analyzed, prototyped, implemented, tuned, and scaled up for leading-edge supercomputers, including UT-Austin’s own 10 petaflops Stampede system, ANL’s Mira system, and ORNL’s Titan system. While our focus is on fundamental mathematical/computational methods and algorithms, we will assess our methods on model problems derived from several DOE mission applications, including multiscale mechanics and ice sheet dynamics.

97 MATHEMATICS AND COMPUTING↗