Search NASA⌕ Search

SEARCH · Search NASA

Results for “Variational inference”

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 181 records · Page 10

Your Clean Graphene is Still Not Clean

Researchers working with thin samples, such as monolayer graphene, are consistently struggling against contamination. Indeed, the problem of hydrocarbon contamination is known from the earliest days of electron microscopy and efforts to reduce this problem are ubiquitous to almost all high-vacuum experiments. Accurate knowledge of the behavior of such contamination is essential for electron beam (e-beam) based atomic fabrication, where it is aspired to select and control matter on an atom-by-atom basis. Here, the vexing question of hydrocarbon contamination on graphene is taken up. Image intensity is used to directly reveal the presence of diffusing hydrocarbons on ostensibly clean graphene. These diffusing hydrocarbons are previously inferred but not directly observed. Surprising dynamic variations of the concentration of these hydrocarbons impels questions about their origin. Here, some possible explanations are presented and some tentative conclusions are drawn. This work updates the conceptual model of “clean graphene” and offers refinements to the description of e-beam induced hydrocarbon deposition.

atomic fabrication↗

In-flight calibration of mesospheric rocket plasma probes

Many effects and factors can influence the efficiency of a rocket plasma probe. These include payload charging, solar illumination, rocket payload orientation and rotation, and dust impact induced secondary charge production. As a consequence, considerable uncertainties can arise in the determination of the effective cross sections of plasma probes and measured electron and ion densities. We present a new method for calibrating mesospheric rocket plasma probes and obtaining reliable measurements of plasma densities. This method can be used if a payload also carries a probe for measuring the dust charge density. It is based on that a dust probe's effective cross section for measuring the charged component of dust normally is nearly equal to its geometric cross section, and it involves the comparison of variations in the dust charge density measured with the dust detector to the corresponding current variations measured with the electron and/or ion probes. In cases in which the dust charge density is significantly smaller than the electron density, the relation between plasma and dust charge density variations can be simplified and used to infer the effective cross sections of the plasma probes. We illustrate the utility of the method by analysing the data from a specific rocket flight of a payload containing both dust and electron probes.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Hydro-scaling of direct-drive cylindrical implosions at the OMEGA and the National Ignition Facility

Deceleration-phase Rayleigh–Taylor instability (RTI) growth during inertial confinement fusion capsule implosions significantly affects the performance as it mixes cold ablator material into the fuel. Precise measurements of such instability growth are essential for both validating the existing simulation codes and improving our predictive capability. RTI measurements on the inner surface of a spherical shell are limited and are often inferred indirectly at limited convergence. In contrast, cylindrical implosions allow for direct diagnostic access to the converging interface by imaging down the cylinder axis while retaining the effects of convergence. We have performed direct-drive cylindrical implosion experiments at both the OMEGA and the NIF laser facilities using scaled targets. RTI growth is demonstrated to be scale-invariant between the cylindrical targets at OMEGA and similar targets at the NIF, which are scaled up by a factor of three in the radial dimension. Single-mode (m = 0) instability growth factors of ~14 are measured at a convergence ratio (CR) ~ 2.5 with nearly identical mode growth at both scales. The measurements are in agreement with xRAGE radiation-hydrodynamics simulations. In addition, we have developed the Bayesian-inference-engine method to account for the variations in the target alignment, magnification, and the parallax effect in the measurement, allowing a more precise comparison between the experimental data and the simulations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

CONSTRAINTS ON THE ASSEMBLY AND DYNAMICS OF GALAXIES. II. PROPERTIES OF KILOPARSEC-SCALE CLUMPS IN REST-FRAME OPTICAL EMISSION OF z {approx} 2 STAR-FORMING GALAXIES

We study the properties of luminous stellar 'clumps' identified in deep, high-resolution Hubble Space Telescope NIC2/F160W imaging at 1.6 {mu}m of six z {approx} 2 star-forming galaxies with existing near-infrared integral field spectroscopy from SINFONI at the Very Large Telescope. Individual clumps contribute {approx}0.5%-15% of the galaxy-integrated rest-frame {approx}5000 A emission, with median of {approx}2%; the total contribution of clump light ranges from 10% to 25%. The median intrinsic clump size and stellar mass are {approx}1 kpc and {approx}10{sup 9} M{sub sun}, in the ranges for clumps identified in rest-UV or line emission in other studies. The clump sizes and masses in the subset of disks are broadly consistent with expectations for clump formation through gravitational instabilities in gas-rich, turbulent disks given the host galaxies' global properties. By combining the NIC2 data with Advanced Camera for Surveys (ACS)/F814W imaging available for one source, and adaptive-optics-assisted SINFONI H{alpha} data for another, we infer modest color, M/L, and stellar age variations within each galaxy. In these two objects, sets of clumps identified at different wavelengths do not fully overlap; NIC2-identified clumps tend to be redder/older than ACS- or H{alpha}-identified clumps without rest-frame optical counterparts. There is evidence for a systematic trend of older ages at smaller galactocentric radii among the clumps, consistent with scenarios where inward migration of clumps transports material toward the central regions. From constraints on a bulge-like component at radii {approx}< 1-3 kpc, none of the five disks in our sample appears to contain a compact massive stellar core, and we do not discern a trend of bulge stellar mass fraction with stellar age of the galaxy. Further observations are necessary to probe the buildup of stellar bulges and the role of clumps in this process.

79 ASTRONOMY AND ASTROPHYSICS↗

INVESTIGATING THE COSMIC-RAY IONIZATION RATE IN THE GALACTIC DIFFUSE INTERSTELLAR MEDIUM THROUGH OBSERVATIONS OF H{sup +}{sub 3}

Observations of H{sup +}{sub 3} in the Galactic diffuse interstellar medium have led to various surprising results, including the conclusion that the cosmic-ray ionization rate ({zeta}{sub 2}) is about one order of magnitude larger than previously thought. The present survey expands the sample of diffuse cloud sight lines with H{sup +}{sub 3} observations to 50, with detections in 21 of those. Ionization rates inferred from these observations are in the range (1.7 {+-} 1.3) Multiplication-Sign 10{sup -16} s{sup -1} < {zeta}{sub 2} < (10.6 {+-} 8.2) Multiplication-Sign 10{sup -16} s{sup -1} with a mean value of {zeta}{sub 2} = (3.5{sup +5.3}{sub -3.0}) Multiplication-Sign 10{sup -16} s{sup -1}. Upper limits (3{sigma}) derived from non-detections of H{sup +}{sub 3} are as low as {zeta}{sub 2} < 0.4 Multiplication-Sign 10{sup -16} s{sup -1}. These low upper limits, in combination with the wide range of inferred cosmic-ray ionization rates, indicate variations in {zeta}{sub 2} between different diffuse cloud sight lines. A study of {zeta}{sub 2} versus N{sub H} (total hydrogen column density) shows that the two parameters are not correlated for diffuse molecular cloud sight lines, but that the ionization rate decreases when N{sub H} increases to values typical of dense molecular clouds. Both the difference in ionization rates between diffuse and dense clouds and the variation of {zeta}{sub 2} among diffuse cloud sight lines are likely the result of particle propagation effects. The lower ionization rate in dense clouds is due to the inability of low-energy (few MeV) protons to penetrate such regions, while the ionization rate in diffuse clouds is controlled by the proximity of the observed cloud to a site of particle acceleration.

79 ASTRONOMY AND ASTROPHYSICS↗

THE BINARITY OF MILKY WAY F,G,K STARS AS A FUNCTION OF EFFECTIVE TEMPERATURE AND METALLICITY

We estimate the fraction of F,G,K stars with close binary companions by analysing multi-epoch stellar spectra from the Sloan Digital Sky Survey (SDSS) and LAMOST for radial velocity variations. We employ a Bayesian method to infer the maximum likelihood of the fraction of binary stars with orbital periods of 1000 days or shorter, assuming a simple model distribution for a binary population with circular orbits. The overall inferred fraction of stars with such a close binary companion is 43.0% ± 2.0% for a sample of F,G,K stars from SDSS SEGUE, and 30% ± 8.0% in a similar sample from LAMOST. The apparent close binary fraction decreases with the stellar effective temperature. We divide the SEGUE and LEGUE data into three subsamples with different metallicity ([Fe/H] < –1.1; –1.1 < [Fe/H] < –0.6; –0.6 < [Fe/H]), for which the inferred close binary fractions are 56 ± 5.0%, 56.0 ± 3%, and 30 ± 5.7%. The metal-rich stars from our sample are therefore substantially less likely to possess a close binary companion than otherwise similar stars drawn from metal-poor populations. The different ages and formation environments of the Milky Way's thin disk, thick disk, and halo may contribute to explaining these observations. Alternatively, metallicity may have a significant effect on the formation and/or evolution of binary stars.

79 ASTRONOMY AND ASTROPHYSICS↗

Equation of State of CO 2 Shock Compressed to 1 TPa

Equation-of-state (pressure, density, temperature, internal energy) and reflectivity measurements on shock-compressed CO 2 at and above the insulating-to-conducting transition reveal new insight into the chemistry of simple molecular systems in the warm-dense-matter regime. CO 2 samples were precompressed in diamond-anvil cells to tune the initial densities from 1.35 g/cm 3 (liquid) to 1.74 g/cm 3 (solid) at room temperature and were then shock compressed up to 1 TPa and 93 000 K. Lastly, variation in initial density was leveraged to infer thermodynamic derivatives including specific heat and Gruneisen coefficient, exposing a complex bonded and moderately ionized state at the most extreme conditions studied.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Performance Study of CXL Memory Topology

This paper presents a comprehensive evaluation of the performance impact of various Compute Express Link (CXL) memory topologies, with a particular emphasis on CXL switches, in the context of High- Performance Computing (HPC) and Large Language Model (LLM) inference workloads. Our study unveils significant performance variations across different topologies, demonstrating that certain configurations yield superior performance for specific workloads. These findings underscore the critical importance of tailored topol- ogy selection in optimizing system performance. Additionally, we address the inherent challenges associated with integrating CXL switches, including overhead considerations and routing complex- ities. Our research highlights the necessity for thorough evalua- tion methodologies to fully leverage CXL technology’s potential in contemporary computing environments. These insights provide valuable guidance for system architects and data center operators in designing and optimizing CXL-based infrastructures for diverse workload requirements.

CXL, memory, Artificial Intelligence (AI), HPC↗

Spatiotemporal Evolution of Hanle and Zeeman Synthetic Polarization in a Chromospheric Spectral Line

Due to the quick evolution of the solar chromosphere, its magnetic field cannot be inferred reliably without accounting for the temporal variations of its polarized light. This has been broadly overlooked in the modeling and interpretation of the polarization, due to technical problems (e.g., lack of temporal resolution or of time-dependent MHD solar models) and/or because many polarization measurements can apparently be explained without dynamics. Here, we show that the temporal evolution is critical for explaining the spectral-line scattering polarization because of its sensitivity to rapidly varying physical quantities and the possibility of signal cancellations and attenuation during extended time integration. For studying the combined effect of time-varying magnetic fields and kinematics, we solved the 1.5D non-LTE problem of the second kind in time-dependent 3D R-MHD solar models and synthesized the Hanle and Zeeman polarization in forward scattering for the chromospheric λ 4227 line. We find that the quiet-Sun polarization amplitudes depend on the periodicity and spectral coherence of the signal enhancements produced by kinematics, but that substantially larger linear polarization signals should exist all over the solar disk for short integration times. The spectral morphology of the polarization is discussed as a combination of Hanle, Zeeman, partial redistribution and dynamic effects. We give physical references for observations by degrading and characterizing our slit time series in different spatiotemporal resolutions. The implications of our results for the interpretation of the second solar spectrum and for the investigation of the solar atmospheric heatings are discussed.

79 ASTRONOMY AND ASTROPHYSICS↗

Bayesian inference of heterogeneous epidemic models: Application to COVID-19 spread accounting for long-term care facilities

Here we propose a high dimensional Bayesian inference framework for learning heterogeneous dynamics of a COVID-19 model, with a specific application to the dynamics and severity of COVID-19 inside and outside long-term care (LTC) facilities. We develop a heterogeneous compartmental model that accounts for the heterogeneity of the time-varying spread and severity of COVID-19 inside and outside LTC facilities, which is characterized by time-dependent stochastic processes and time-independent parameters in ~ 1500 dimensions after discretization. To infer these parameters, we use reported data on the number of confirmed, hospitalized, and deceased cases with suitable post-processing in both a deterministic inversion approach with appropriate regularization as a first step, followed by Bayesian inversion with proper prior distributions. To address the curse of dimensionality and the ill-posedness of the high-dimensional inference problem, we propose use of a dimension-independent projected Stein variational gradient descent method, and demonstrate the intrinsic low-dimensionality of the inverse problem. We present inference results with quantified uncertainties for both New Jersey and Texas, which experienced different epidemic phases and patterns. Moreover, we also present forecasting and validation results based on the empirical posterior samples of our inference for the future trajectory of COVID-19.

42 ENGINEERING↗

Brittleness and Bayesian Inference

In a recent report, “Bayesian Brittleness: Why no Bayesian model is ‘good enough,’” (arXiv, 1304.6772v1) the authors prove rigorously that under a seemingly natural class of priors Π, the posterior expectation of a function Φ(θ) can achieve essentially any value that Φ(θ) alone can achieve, so that Bayesian inference appears to have no robustness whatsoever. We explain this puzzling result, and show why it does not imply a breakdown of Bayesian inference. The problem is that the priors leading to the extreme results depend on the observed data, which means they are not valid priors. The corresponding posteriors are also invalid, so the extreme variation in these posteriors does not imply that Bayesian inference is brittle. The report, however, provides a new framework which may be used for any assumed class of priors Π, and which we explain in detail. This framework may be a useful new tool for deriving quantitative robustness results.

97 MATHEMATICS AND COMPUTING↗

Ensemble variational Fokker-Planck methods for data assimilation

Particle flow filters solve Bayesian inference problems by smoothly transforming a set of particles into samples from the posterior distribution. Particles move in state space under the flow of an McKean-Vlasov-Itˆo process. This work introduces the Variational Fokker-Planck (VFP) framework for data assimilation, a general approach that includes previously known particle flow filters as special cases. The McKean-Vlasov-Itˆo process that transforms particles is defined via an optimal drift that depends on the selected diffusion term. It is established that the underlying probability density - sampled by the ensemble of particles - converges to the Bayesian posterior probability density. For a finite number of particles the optimal drift contains a regularization term that nudges particles toward becoming independent random variables. Based on this analysis, we derive computationally-feasible approximate regularization approaches that penalize the mutual information between pairs of particles, and avoid particle collapse. Moreover, the diffusion plays a role akin to a particle rejuvenation approach that aims to alleviate particle collapse. The VFP framework is very flexible. Different assumptions on prior and intermediate probability distributions can be used to implement the optimal drift, and localization and covariance shrinkage can be applied to alleviate the curse of dimensionality. A robust implicit-explicit method is discussed for the efficient integration of stiff McKean- Vlasov-Itˆo processes. Here, the effectiveness of the VFP framework is demonstrated on three progressively more challenging test problems, namely the Lorenz ’63, Lorenz ’96 and the quasi-geostrophic equations.

97 MATHEMATICS AND COMPUTING↗

Variational Bayesian strategies for high-dimensional, stochastic design problems

This paper is concerned with a lesser-studied problem in the context of model-based, uncertainty quantification (UQ), that of optimization/design/control under uncertainty. The solution of such problems is hindered not only by the usual difficulties encountered in UQ tasks (e.g. the high computational cost of each forward simulation, the large number of random variables) but also by the need to solve a nonlinear optimization problem involving large numbers of design variables and potentially constraints. We propose a framework that is suitable for a class of such problems and is based on the idea of recasting them as probabilistic inference tasks. To that end, we propose a Variational Bayesian (VB) formulation and an iterative VB–Expectation-Maximization scheme that is capable of identifying a local maximum as well as a low-dimensional set of directions in the design space, along which, the objective exhibits the largest sensitivity. We demonstrate the validity of the proposed approach in the context of two numerical examples involving thousands of random and design variables. In all cases considered the cost of the computations in terms of calls to the forward model was of the order of 100 or less. The accuracy of the approximations provided is assessed by information-theoretic metrics.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Inferring demographic and selective histories from population genomic data using a 2-step approach in species with coding-sparse genomes: an application to human data

Abstract The demographic history of a population, and the distribution of fitness effects (DFE) of newly arising mutations in functional genomic regions, are fundamental factors dictating both genetic variation and evolutionary trajectories. Although both demographic and DFE inference has been performed extensively in humans, these approaches have generally either been limited to simple demographic models involving a single population, or, where a complex population history has been inferred, without accounting for the potentially confounding effects of selection at linked sites. Taking advantage of the coding-sparse nature of the genome, we propose a 2-step approach in which coalescent simulations are first used to infer a complex multi-population demographic model, utilizing large non-functional regions that are likely free from the effects of background selection. We then use forward-in-time simulations to perform DFE inference in functional regions, conditional on the complex demography inferred and utilizing expected background selection effects in the estimation procedure. Throughout, recombination and mutation rate maps were used to account for the underlying empirical rate heterogeneity across the human genome. Importantly, within this framework it is possible to utilize and fit multiple aspects of the data, and this inference scheme represents a generalized approach for such large-scale inference in species with coding-sparse genomes.

Soni, Vivak (ORCID:0000000294969562)↗

TIME EVOLUTION OF PLASMA PARAMETERS DURING THE RISE OF A SOLAR PROMINENCE INSTABILITY

We present high-spatial resolution spectropolarimetric observations of a quiescent hedgerow prominence taken in the He I 1083.0 nm triplet. The observation consisted of a time series in sit-and-stare mode of ∼36 minutes duration. The spectrograph's slit crossed the prominence body and we recorded the time evolution of individual vertical threads. Eventually, we observed the development of a dark Rayleigh-Taylor plume that propagated upward with a velocity, projected onto the plane of the sky, of 17 km s{sup –1}. Interestingly, the plume apex collided with the prominence threads pushing them aside. We inferred Doppler shifts, Doppler widths, and magnetic field strength variations by interpreting the He I Stokes profiles with the HAZEL code. The Doppler shifts show that clusters of threads move coherently while individual threads have oscillatory patterns. Regarding the plume we found strong redshifts (∼9-12 km s{sup –1}) and large Doppler widths (∼10 km s{sup –1}) at the plume apex when it passed through the prominence body and before it disintegrated. We associate the redshifts with perspective effects while the Doppler widths are more likely due to an increase in the local temperature. No local variations of the magnetic field strength associated with the passage of the plume were found; this leads us to conclude that the plumes are no more magnetized than the surroundings. Finally, we found that some of the threads' oscillations are locally damped, what allowed us to apply prominence seismology techniques to infer additional prominence physical parameters.

79 ASTRONOMY AND ASTROPHYSICS↗

Uberon, an integrative multi-species anatomy ontology

We present Uberon, an integrated cross-species ontology consisting of over 6,500 classes representing a variety of anatomical entities, organized according to traditional anatomical classification criteria. The ontology represents structures in a species-neutral way and includes extensive associations to existing species-centric anatomical ontologies, allowing integration of model organism and human data. Uberon provides a necessary bridge between anatomical structures in different taxa for cross-species inference. It uses novel methods for representing taxonomic variation, and has proved to be essential for translational phenotype analyses. Uberon is available at http://uberon.org

59 BASIC BIOLOGICAL SCIENCES↗

MetaPop: a pipeline for macro- and microdiversity analyses and visualization of microbial and viral metagenome-derived populations

Abstract Background Microbes and their viruses are hidden engines driving Earth’s ecosystems from the oceans and soils to humans and bioreactors. Though gene marker approaches can now be complemented by genome-resolved studies of inter-(macrodiversity) and intra-(microdiversity) population variation, analytical tools to do so remain scattered or under-developed. Results Here, we introduce MetaPop, an open-source bioinformatic pipeline that provides a single interface to analyze and visualize microbial and viral community metagenomes at both the macro - and microdiversity levels. Macrodiversity estimates include population abundances and α- and β-diversity. Microdiversity calculations include identification of single nucleotide polymorphisms, novel codon-constrained linkage of SNPs, nucleotide diversity ( π and θ ), and selective pressures (pN/pS and Tajima’s D ) within and fixation indices ( F ST ) between populations. MetaPop will also identify genes with distinct codon usage. Following rigorous validation, we applied MetaPop to the gut viromes of autistic children that underwent fecal microbiota transfers and their neurotypical peers. The macrodiversity results confirmed our prior findings for viral populations (microbial shotgun metagenomes were not available) that diversity did not significantly differ between autistic and neurotypical children. However, by also quantifying microdiversity, MetaPop revealed lower average viral nucleotide diversity ( π ) in autistic children. Analysis of the percentage of genomes detected under positive selection was also lower among autistic children, suggesting that higher viral π in neurotypical children may be beneficial because it allows populations to better “bet hedge” in changing environments. Further, comparisons of microdiversity pre- and post-FMT in autistic children revealed that the delivery FMT method (oral versus rectal) may influence viral activity and engraftment of microdiverse viral populations, with children who received their FMT rectally having higher microdiversity post-FMT. Overall, these results show that analyses at the macro level alone can miss important biological differences. Conclusions These findings suggest that standardized population and genetic variation analyses will be invaluable for maximizing biological inference, and MetaPop provides a convenient tool package to explore the dual impact of macro - and microdiversity across microbial communities.

59 BASIC BIOLOGICAL SCIENCES↗

Signatures of optimal codon usage in metabolic genes inform budding yeast ecology

Reverse ecology is the inference of ecological information from patterns of genomic variation. One rich, heretofore underutilized, source of ecologically relevant genomic information is codon optimality or adaptation. Bias toward codons that match the tRNA pool is robustly associated with high gene expression in diverse organisms, suggesting that codon optimization could be used in a reverse ecology framework to identify highly expressed, ecologically relevant genes. To test this hypothesis, we examined the relationship between optimal codon usage in the classic galactose metabolism ( GAL ) pathway and known ecological niches for 329 species of budding yeasts, a diverse subphylum of fungi. We find that optimal codon usage in the GAL pathway is positively correlated with quantitative growth on galactose, suggesting that GAL codon optimization reflects increased capacity to grow on galactose. Optimal codon usage in the GAL pathway is also positively correlated with human-associated ecological niches in yeasts of the CUG-Ser1 clade and with dairy-associated ecological niches in the family Saccharomycetaceae. For example, optimal codon usage of GAL genes is greater than 85% of all genes in the genome of the major human pathogen Candida albicans (CUG-Ser1 clade) and greater than 75% of genes in the genome of the dairy yeast Kluyveromyces lactis (family Saccharomycetaceae). We further find a correlation between optimization in the GAL actose pathway genes and several genes associated with nutrient sensing and metabolism. This work suggests that codon optimization harbors information about the metabolic ecology of microbial eukaryotes. This information may be particularly useful for studying fungal dark matter—species that have yet to be cultured in the lab or have only been identified by genomic material.

59 BASIC BIOLOGICAL SCIENCES↗