Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Variational Bayesian”

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 199 records · Page 11

Model reduction methods for nuclear emulators

The field of model order reduction (MOR) is growing in importance due to its ability to extract the key insights from complex simulations while discarding computationally burdensome and superfluous information. We provide an overview of MOR methods for the creation of fast & accurate emulators of memory- and compute-intensive nuclear systems, focusing on eigen-emulators and variational emulators. As an example, we describe how 'eigenvector continuation' is a special case of a much more general and well-studied MOR formalism for parameterized systems. We continue with an introduction to the Ritz and Galerkin projection methods that underpin many such emulators, while pointing to the relevant MOR theory and its successful applications along the way. Here, we believe that this guide will open the door to broader applications in nuclear physics and facilitate communication with practitioners in other fields.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

NASA's Human Mission to a Near-Earth Asteroid: Landing on a Moving Target

This paper describes a Bayesian approach for comparing the productivity and cost-risk tradeoffs of sending versus not sending one or more robotic surveyor missions prior to a human mission to land on an asteroid. The expected value of sample information based on productivity combined with parametric variations in the prior probability an asteroid might be found suitable for landing were used to assess the optimal number of spacecraft and asteroids to survey. The analysis supports the value of surveyor missions to asteroids and indicates one launch with two spacecraft going simultaneously to two independent asteroids appears optimal.

precursor decision making↗

Extending Parsimonious Bayesian Inference

Parsimonious Bayesian inference is a theoretical framework for efficient data assimilation that seeks to balance increased consistency between predictions and training data against corresponding increases in model complexity. Within this framework, over-training is understood as optimization that encodes excessive information within model parameters while only achieving small improvements between predictions and training data. This project aims to develop practical methods of limiting excess model information during optimization. One key observation is that practical heuristics for parsimonious learning in high-dimensions must balance expressivity, i.e. the ability of the model to capture diverse predictions with only a few non-zero parameters, against discoverability, i.e. the ability to train the model with gradient-based optimization and drive parameters to low information states. As such, we developed logical activation functions that are able to adaptively approximate arbitrary truth tables that define Boolean logic operations within a probabilistic framework. These functions have demonstrated the ability to learn exclusive disjunction (XOR) and conditioned disjunction (if [condition] then [result_if_true] else [result_if_false]) within a single layer of a neural network. To efficiently exploit these activation functions to drive parsimonious learning required several other advances within the domain of variational inference. The most efficient form of complexity suppression is structured sparsification, driving most model parameters to zero while achieving the structural coherence among nonzeros needed for bandwidth reduction. Such models are not only far more efficient at suppressing information-theoretic complexity, they also reduce the other forms of complexity (computations, communication, storage, and the number of dependencies needed to evaluate predictions). Aiming to support enhanced sparsification, this project examined new approaches to high-dimensional variational inference that allow us to calibrate and control parameter uncertainty during optimization. By identifying which parameters can sustain sparsifying perturbations with little impact on prediction quality, we can develop better pruning strategies by framing them as approximate Bayesian inference. These advances also open paths to mitigate concerns with deploying advanced learning methods in resource-constrained environments, such as running models on power-limited or communication-limited devices.

97 MATHEMATICS AND COMPUTING↗

Goal-oriented a-posteriori estimation of model error as an aid to parameter estimation

In this work, a Bayesian model calibration framework is presented that utilizes goal-oriented a-posterior error estimates in quantities of interest (QoIs) for classes of high-fidelity models characterized by PDEs. It is shown that for a large class of computational models, it is possible to develop a computationally inexpensive procedure for calibrating parameters of high-fidelity models of physical events when the parameters of low-fidelity (surrogate) models are known with acceptable accuracy. The main ingredients in the proposed model calibration scheme are goal-oriented a-posteriori estimates of error in QoIs computed using a so-called lower fidelity model compared to those of an uncalibrated higher fidelity model. The estimates of error in QoIs are used to define likelihood functions in Bayesian inversion analysis. A standard Bayesian approach is employed to compute the posterior distribution of model parameters of high-fidelity models. As applications, parameters in a quasi-linear second-order elliptic boundary-value problem (BVP) are calibrated using a second-order linear elliptic BVP. In a second application, parameters of a tumor growth model involving nonlinear time-dependent PDEs are calibrated using a lower fidelity linear tumor growth model with known parameter values.

A-posterior estimates↗

CODEX weak lensing mass catalogue and implications on the mass–richness relation

The COnstrain Dark Energy with X-ray clusters (CODEX) sample contains the largest flux limited sample of X-ray clusters at 0.35 < z < 0.65. It was selected from ROSAT data in the 10 000 square degrees of overlap with BOSS, mapping a total number of 2770 high-z galaxy clusters. We present here the full results of the CFHT CODEX programme on cluster mass measurement, including a reanalysis of CFHTLS Wide data, with 25 individual lensing-constrained cluster masses. We employ lensfit shape measurement and perform a conservative colour–space selection and weighting of background galaxies. Using the combination of shape noise and an analytic covariance for intrinsic variations of cluster profiles at fixed mass due to large-scale structure, miscentring, and variations in concentration and ellipticity, we determine the likelihood of the observed shear signal as a function of true mass for each cluster. We combine 25 individual cluster mass likelihoods in a Bayesian hierarchical scheme with the inclusion of optical and X-ray selection functions to derive constraints on the slope α, normalization β, and scatter σln λ|μ of our richness–mass scaling relation model in log-space: ${\langle {\rm In}\,\, \lambda\!\!\mid\!\!\mu\rangle = \alpha\mu + \beta,} $ with μ = ln (M 200c /M piv ), and M piv = 10 14.81 M ⊙ . We find a slope $\alpha = 0.49^{+0.20}_{-0.15}$, normalization $\exp (\beta) = 84.0^{+9.2}_{-14.8}$, and $\sigma _{\ln \lambda | \mu } = 0.17^{+0.13}_{-0.09}$ using CFHT richness estimates. In comparison to other weak lensing richness–mass relations, we find the normalization of the richness statistically agreeing with the normalization of other scaling relations from a broad redshift range (0.0 < z < 0.65) and with different cluster selection (X-ray, Sunyaev–Zeldovich, and optical).

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

PNNL-CompBio/bayesian-inference

Bayesian Metabolic Inference is a platform to integrate genome-scale multi-omics data into an iterative method to obtain optimized recommendations for optimal target compound production. This method uses a variational inference approach to approximate new posterior data to analyze in order to obtain useful correlations and control coefficients.

Kumar, Neeraj↗

Probing Blazar Emission Processes with Optical/Gamma-Ray Flare Correlations

We report that even with several thousand Fermi-Large Area Telescope (LAT) blazar detections, the γ-ray emission mechanism is poorly understood. We explore correlated optical/γ-ray flux variations for 178 Fermi-LAT blazars regularly monitored by the Katzman Automatic Imaging Telescope, the Small and Moderate Aperture Research Telescope System, and the Steward Observatory. Out of the 178 sources, 121 show a measurable (>1σ) discrete correlation function peak. Using the derived time lags and Bayesian block light-curve decompositions, we measure the fraction of common and orphan flares between the two bands. After accounting for sampling and sensitivity limitations we quantify for the first time the true orphan flare rates of optical and γ-ray flares: 54.5% of optical and 20% of γ-ray flares are orphan events. Both the intraband temporal relation and the small orphan γ-ray flare fraction point toward leptonic processes as the likely mechanisms for the high-energy emission. Motivated to discriminate between synchrotron self-Compton (SSC) and external-Compton dominance in individual sources, we use the flux–flux variations to determine the slope m of the log f opt –log f γ dependence. The slope distribution suggests a bimodal population with high and intermediate synchrotron peak objects showing larger m than low synchrotron peak objects. We find that m is naturally decreased through pollution from the orphan (typically optical) flares and develop a method to statistically recover, given the sources' measured orphan flare rate, the intrinsic m. Finally, while source classes show composite behavior, the majority of BL Lac objects favor m = 2, indicating an SSC origin for the γ-rays. No preference for either m is found in flat spectrum radio quasars.

79 ASTRONOMY AND ASTROPHYSICS↗

General-Purpose Bayesian Tensor Learning With Automatic Rank Determination and Uncertainty Quantification

A major challenge in many machine learning tasks is that the model expressive power depends on model size. Low-rank tensor methods are an efficient tool for handling the curse of dimensionality in many large-scale machine learning models. The major challenges in training a tensor learning model include how to process the high-volume data, how to determine the tensor rank automatically, and how to estimate the uncertainty of the results. While existing tensor learning focuses on a specific task, this paper proposes a generic Bayesian framework that can be employed to solve a broad class of tensor learning problems such as tensor completion, tensor regression, and tensorized neural networks. We develop a low-rank tensor prior for automatic rank determination in nonlinear problems. Our method is implemented with both stochastic gradient Hamiltonian Monte Carlo (SGHMC) and Stein Variational Gradient Descent (SVGD). We compare the automatic rank determination and uncertainty quantification of these two solvers. We demonstrate that our proposed method can determine the tensor rank automatically and can quantify the uncertainty of the obtained results. We validate our framework on tensor completion tasks and tensorized neural network training tasks.

Bayesian inference↗

Regional Sources of Nitrous Oxide over the United States: Seasonal Variation and Spatial Distribution

This paper presents top-down constraints on the magnitude, spatial distribution, and seasonality of nitrous oxide (N{sub 2}O) emissions over the central United States. We analyze data from tall towers in 2004 and 2008 using a high resolution Lagrangian particle dispersion model paired with both geostatistical and Bayesian inversions. Our results indicate peak N{sub 2}O emissions in June with a strong seasonal cycle. The spatial distribution of sources closely mirrors data on fertilizer application with particularly large N{sub 2}O sources over the US Cornbelt. Existing inventories for N{sub 2}O predict emissions that differ substantially from the inverse model results in both seasonal cycle and magnitude. We estimate a total annual N{sub 2}O budget over the central US of 0.9-1.2 TgN/yr and an extrapolated budget for the entire US and Canada of 2.1-2.6 TgN/yr. By this estimate, the US and Canada account for 12-15% of the total global N{sub 2}O source or 32-39% of the global anthropogenic source as reported by the Intergovernmental Panel on Climate Change in 2007.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Probabilistic Nanomagnetic Memories for Uncertain and Robust Machine Learning

This project evaluated the use of emerging spintronic memory devices for robust and efficient variational inference schemes. Variational inference (VI) schemes, which constrain the distribution for each weight to be a Gaussian distribution with a mean and standard deviation, are a tractable method for calculating posterior distributions of weights in a Bayesian neural network such that this neural network can also be trained using the powerful backpropagation algorithm. Our project focuses on domain-wall magnetic tunnel junctions (DW-MTJs), a powerful multi-functional spintronic synapse design that can achieve low power switching while also opening the pathway towards repeatable, analog operation using fabricated notches. Our initial efforts to employ DW-MTJs as an all-in-one stochastic synapse with both a mean and standard deviation didn’t end up meeting the quality metrics for hardware-friendly VI. In the future, new device stacks and methods for expressive anisotropy modification may make this idea still possible. However, as a fall back that immediately satisfies our requirements, we invented and detailed how the combination of a DW-MTJ synapse encoding the mean and a probabilistic Bayes-MTJ device, programmed via a ferroelectric or ionically modifiable layer, can robustly and expressively implement VI. This design includes a physics-informed small circuit model, that was scaled up to perform and demonstrate rigorous uncertainty quantification applications, up to and including small convolutional networks on a grayscale image classification task, and larger (Residual) networks implementing multi-channel image classification. Lastly, as these results and ideas all depend upon the idea of an inference application where weights (spintronic memory states) remain non-volatile, the retention of these synapses for the notched case was further interrogated. These investigations revealed and emphasized the importance of both notch geometry and anisotropy modification in order to further enhance the endurance of written spintronic states. In the near future, these results will be mapped to effective predictions for room temperature and elevated operation DW-MTJ memory retention, and experimentally verified when devices become available.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Parametric Analysis of a Hover Test Vehicle using Advanced Test Generation and Data Analysis

Large complex aerospace systems are generally validated in regions local to anticipated operating points rather than through characterization of the entire feasible operational envelope of the system. This is due to the large parameter space, and complex, highly coupled nonlinear nature of the different systems that contribute to the performance of the aerospace system. We have addressed the factors deterring such an analysis by applying a combination of technologies to the area of flight envelop assessment. We utilize n-factor (2,3) combinatorial parameter variations to limit the number of cases, but still explore important interactions in the parameter space in a systematic fashion. The data generated is automatically analyzed through a combination of unsupervised learning using a Bayesian multivariate clustering technique (AutoBayes) and supervised learning of critical parameter ranges using the machine-learning tool TAR3, a treatment learner. Covariance analysis with scatter plots and likelihood contours are used to visualize correlations between simulation parameters and simulation results, a task that requires tool support, especially for large and complex models. We present results of simulation experiments for a cold-gas-powered hover test vehicle.

Gundy-Burlet, Karen↗

Bayesian learning of orthogonal embeddings for multi-fidelity Gaussian Processes

Uncertainty propagation in complex engineering systems often poses significant computational challenges related to modeling and quantifying probability distributions of model outputs, as those emerge as the result of various sources of uncertainty that are inherent in the system under investigation. Gaussian Processes regression (GPs) is a robust meta-modeling technique that allows for fast model prediction and exploration of response surfaces. Multi-fidelity variations of GPs further leverage information from cheap and low fidelity model simulations in order to improve their predictive performance on the high fidelity model. In order to cope with the high volume of data required to train GPs in high dimensional design spaces, a common practice is to introduce latent design variables that are typically projections of the original input space to a lower dimensional subspace, and therefore substitute the problem of learning the initial high dimensional mapping, with that of training a GP on a low dimensional space. Here in this paper, we present a Bayesian approach to identify optimal transformations that map the input points to low dimensional latent variables. The \projection" mapping consists of an orthonormal matrix that is considered a priori unknown and needs to be inferred jointly with the GP parameters, conditioned on the available training data. The proposed Bayesian inference scheme relies on a two-step iterative algorithm that samples from the marginal posteriors of the GP parameters and the projection matrix respectively, both using Markov Chain Monte Carlo (MCMC) sampling. In order to take into account the orthogonality constraints imposed on the orthonormal projection matrix, a Geodesic Monte Carlo sampling algorithm is employed, that is suitable for exploiting probability measures on manifolds. We extend the proposed framework to multi-fidelity models using GPs including the scenarios of training multiple outputs together. We validate our framework on three synthetic problems with a known lower-dimensional subspace. The benefits of our proposed framework, are illustrated on the computationally challenging aerodynamic optimization of a last-stage blade for an industrial gas turbine, where we study the effect of an 85-dimensional shape parameterization of a three-dimensional airfoil on two output quantities of interest, specifically on the aerodynamic efficiency and the degree of reaction

42 ENGINEERING↗

Bayesian Abel Inversion in Quantitative X-Ray Radiography

A common image formation process in high-energy X-ray radiography is to have a pulsed power source that emits X-rays through a scene, a scintillator that absorbs X-rays and uoresces in the visible spectrum in response to the absorbed photons, and a CCD camera that images the visible light emitted from the scintillator. The intensity image is related to areal density, and, for an object that is radially symmetric about a central axis, the Abel transform then gives the object's volumetric density. Two of the primary drawbacks to classical variational methods for Abel inversion are their sensitivity to the type and scale of regularization chosen and the lack of natural methods for quantifying the uncertainties associated with the reconstructions. In this work we cast the Abel inversion problem within a statistical framework in order to compute volumetric object densities from X-ray radiographs and to quantify uncertainties in the reconstruction. A hierarchical Bayesian model is developed with a likelihood based on a Gaussian noise model and with priors placed on the unknown density pro le, the data precision matrix, and two scale parameters. This allows the data to drive the localization of features in the reconstruction and results in a joint posterior distribution for the unknown density pro le, the prior parameters, and the spatial structure of the precision matrix. Results of the density reconstructions and pointwise uncertainty estimates are presented for both synthetic signals and real data from a U.S. Department of Energy X-ray imaging facility.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Mapping Aedes aegypti (Diptera: Culicidae) and Aedes albopictus Vector Mosquito Distribution in Brownsville, TX

Abstract Aedes mosquitoes are vectors of several emerging diseases and are spreading worldwide. We investigated the spatiotemporal dynamics of Aedes aegypti (Linnaeus) and Aedes albopictus (Skuse) mosquito trap captures in Brownsville, TX, using high-resolution land cover, socioeconomic, and meteorological data. We modeled mosquito trap counts using a Bayesian hierarchical mixed-effects model with spatially correlated residuals. The models indicated an inverse relationship between temperature and mosquito trap counts for both species, which may be due to the hot and arid climate of southern Texas. The temporal trend in mosquito populations indicated Ae. aegypti populations peaking in the late spring and Ae. albopictus reaching a maximum in winter. Our results indicated that seasonal weather variation, vegetation height, human population, and land cover determine which of the two Aedes species will predominate.

59 BASIC BIOLOGICAL SCIENCES↗

Uncertainty Quantification of Multi-Phase Closures

In the ensemble-averaged dispersed phase formulation used for CFD of multiphase ows in nuclear reactor thermohydraulics, closures of interphase transfer of mass, momentum, and energy constitute, by far, the biggest source of error and uncertainty. Reliable estimators of this source of error and uncertainty are currently non-existent. Here, we report on how modern Validation and Uncertainty Quanti cation (VUQ) techniques can be leveraged to not only quantify such errors and uncertainties, but also to uncover (unintended) interactions between closures of di erent phenomena. As such this approach serves as a valuable aide in the research and development of multiphase closures. The joint modeling of lift, drag, wall lubrication, and turbulent dispersion|forces that lead to tranfer of momentum between the liquid and gas phases|is examined in the frame- work of validation of the adiabatic but turbulent experiments of Liu and Banko , 1993. An extensive calibration study is undertaken with a popular combination of closure relations and the popular k-ϵ turbulence model in a Bayesian framework. When a wide range of super cial liquid and gas velocities and void fractions is considered, it is found that this set of closures can be validated against the experimental data only by allowing large variations in the coe cients associated with the closures. We argue that such an extent of variation is a measure of uncertainty induced by the chosen set of closures. We also nd that while mean uid velocity and void fraction pro les are properly t, uctuating uid velocity may or may not be properly t. This aspect needs to be investigated further. The popular set of closures considered contains ad-hoc components and are undesirable from a predictive modeling point of view. Consequently, we next consider improvements that are being developed by the MIT group under CASL and which remove the ad-hoc elements. We use non-intrusive methodologies for sensitivity analysis and calibration (using Dakota) to study sensitivities of the CFD representation (STARCCM+) of uid velocity pro les and void fraction pro les in the context of Shaver and Podowski, 2015 correction to lift, and the Lubchenko et al., 2017 formulation of wall lubrication.

97 MATHEMATICS AND COMPUTING↗

New constraints on sodium production in globular clusters from the Na 23 ( He 3 , d ) Mg 24 reaction

The star-to-star anticorrelation of sodium and oxygen is a defining feature of globular clusters, but, to date, the astrophysical site responsible for this unique chemical signature remains unknown. Sodium enrichment within these clusters depends sensitively on reaction rate of the sodium destroying reactions 23 Na(p, γ) and 23 Na(p,α). In this paper, we report the results of a 23 Na( 3 He,d) 24 Mg transfer reaction carried out at Triangle Universities Nuclear Laboratory using a 21 MeV 3 He beam. Astrophysically relevant states in 24 Mg between 11 < E x < 12 MeV were studied using high-resolution magnetic spectroscopy, thereby allowing the extraction of excitation energies and spectroscopic factors. Bayesian methods are combined with the distorted wave Born approximation to assign statistically meaningful uncertainties to the extracted spectroscopic factors. For the first time, these uncertainties are propagated through to the estimation of proton partial widths. Our experimental data are used to calculate the reaction rate. The impact of the new rates are investigated using asymptotic giant branch star models. Furthermore, it is found that while the astrophysical conditions still dominate the total uncertainty, intramodel variations on sodium production from the 23 Na(p, γ) and 23 Na(p,α) reaction channels are a lingering source of uncertainty.

20 ≤ A ≤ 38↗

New analysis indicates no thermal inversion in the atmosphere of HD 209458b

An important focus of exoplanet research is the determination of the atmospheric temperature structure of strongly irradiated gas giant planets, or hot Jupiters. HD 209458b is the prototypical exoplanet for atmospheric thermal inversions, but this assertion does not take into account recently obtained data or newer data reduction techniques. We reexamine this claim by investigating all publicly available Spitzer Space Telescope secondary-eclipse photometric data of HD 209458b and performing a self-consistent analysis. We employ data reduction techniques that minimize stellar centroid variations, apply sophisticated models to known Spitzer systematics, and account for time-correlated noise in the data. We derive new secondary-eclipse depths of 0.119% ± 0.007%, 0.123% ± 0.006%, 0.134% ± 0.035%, and 0.215% ± 0.008% in the 3.6, 4.5, 5.8, and 8.0 μm bandpasses, respectively. We feed these results into a Bayesian atmospheric retrieval analysis and determine that it is unnecessary to invoke a thermal inversion to explain our secondary-eclipse depths. The data are well fitted by a temperature model that decreases monotonically between pressure levels of 1 and 0.01 bars. We conclude that there is no evidence for a thermal inversion in the atmosphere of HD 209458b.

79 ASTRONOMY AND ASTROPHYSICS↗