Search NASA⌕ Search

SEARCH · Search NASA

Results for “Bayesian inversion modeling”

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 307 records · Page 17

Uncertainty estimates for the Bayes Inference Engine, (BIE)

In the fall 2007 meeting of the BIB users group, two approaches to making uncertainty estimates were presented. Ken Hanson asserted that if the BFGS optimizer was used, the inverse Hessian matrix was the same as the covariance matrix representing parameter uncertainties. John Pang presented preliminary results of a Monte Carlo method called Randomized Maximum Likelihood (RML). The BFGS/Hessian matrix approach may be applied to the region of the 'ideal model' Approximately 250 parameters describing the object density patches that are varied to match an image of 1,000,000 pixels. I cast this in terms of least squares analysis, as it is much better understood. This not as large a conceptual jump as some suppose because many of the functional blocks in the BIB are taken directly from existing least squares programs. If a Gaussian (normal) probability density function is assumed for both the observation and parameter errors, the Bayesian and least squares result should be identical.

97 MATHEMATICS AND COMPUTING↗

Fast alogorithms for Bayesian uncertainty quantification in large-scale linear inverse problems based on low-rank partial Hessian approximations

We consider the problem of estimating the uncertainty in large-scale linear statistical inverse problems with high-dimensional parameter spaces within the framework of Bayesian inference. When the noise and prior probability densities are Gaussian, the solution to the inverse problem is also Gaussian, and is thus characterized by the mean and covariance matrix of the posterior probability density. Unfortunately, explicitly computing the posterior covariance matrix requires as many forward solutions as there are parameters, and is thus prohibitive when the forward problem is expensive and the parameter dimension is large. However, for many ill-posed inverse problems, the Hessian matrix of the data misfit term has a spectrum that collapses rapidly to zero. We present a fast method for computation of an approximation to the posterior covariance that exploits the lowrank structure of the preconditioned (by the prior covariance) Hessian of the data misfit. Analysis of an infinite-dimensional model convection-diffusion problem, and numerical experiments on large-scale 3D convection-diffusion inverse problems with up to 1.5 million parameters, demonstrate that the number of forward PDE solves required for an accurate low-rank approximation is independent of the problem dimension. This permits scalable estimation of the uncertainty in large-scale ill-posed linear inverse problems at a small multiple (independent of the problem dimension) of the cost of solving the forward problem.

97 MATHEMATICS AND COMPUTING↗

Point Spread Function Estimation and Uncertainty Quantification

An important component of analyzing images quantitatively is modeling image blur due to effects from the system for image capture. When the effect of image blur is assumed to be translation invariant and isotropic, it can be generally modeled as convolution with a radially symmetric kernel, called the point spread function (PSF). Standard techniques for estimating the PSF involve imaging a bright point source, but this is not always feasible (e.g. high energy radiography). This work provides a novel non-parametric approach to estimating the PSF from a calibration image of a vertical edge. Moreover, the approach is within a hierarchical Bayesian framework that in addition to providing a method for estimation, also gives a quantification of uncertainty in the estimate by Markov Chain Monte Carlo methods.

97 MATHEMATICS AND COMPUTING↗

Cosmic Inference: Constraining Parameters with Observations and a Highly Limited Number of Simulations

We look at cosmological probes that pose an inverse problem where the measurement result is obtained through observations, and the objective is to infer values of model parameters that characterize the underlying physical system-our universe, from these observations and theoretical forward-modeling. The only way to accurately forward-model physical behavior on small scales is via expensive numerical simulations, which are further “emulated” due to their high cost. Emulators are commonly built with a set of simulations covering the parameter space with Latin hypercube sampling and an interpolation procedure; the aim is to establish an approximately constant prediction error across the hypercube. In this paper, we provide a description of a novel statistical framework for obtaining accurate parameter constraints. The proposed framework uses multi-output Gaussian process emulators that are adaptively constructed using Bayesian optimization methods with the goal of maintaining a low emulation error in the region of the hypercube preferred by the observational data. In this paper, we compare several approaches for constructing multi-output emulators that enable us to take possible inter-output correlations into account while maintaining the efficiency needed for inference. Using a Lyα forest flux power spectrum, we demonstrate that our adaptive approach requires considerably fewer-by a factor of a few in the Lyα P(k) case considered here-simulations compared to the emulation based on Latin hypercube sampling, and that the method is more robust in reconstructing parameters and their Bayesian credible intervals.

79 ASTRONOMY AND ASTROPHYSICS↗

Summary report of the 4th IAEA Technical Meeting on Fusion Data Processing, Validation and Analysis (FDPVA)

The objective of the Fourth Technical Meeting on Fusion Data Processing, Validation and Analysis was to provide a platform during which a set of topics relevant to fusion data processing, validation and analysis are discussed with the view of extrapolating needs to next step fusion devices such as ITER. The validation and analysis of experimental data obtained from diagnostics used to characterize fusion plasmas are crucial for a knowledge-based understanding of the physical processes governing the dynamics of these plasmas. This paper presents the recent progress and achievements in the domain of plasma diagnostics and synthetic diagnostics data analysis (including image processing, regression analysis, inverse problems, deep learning, machine learning, big data and physics-based models for control) reported at the meeting. The progress in these areas highlight trends observed in current major fusion confinement devices. A special focus is dedicated on data analysis requirements for ITER and DEMO with a particular attention paid to artificial intelligence for automatization and improving reliability of control processes.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

DETECTING RELATIVISTIC X-RAY JETS IN HIGH-REDSHIFT QUASARS

We analyze Chandra X-ray images of a sample of 11 quasars that are known to contain kiloparsec scale radio jets. The sample consists of five high-redshift ( z ≥ 3.6) flat-spectrum radio quasars, and six intermediate redshift (2.1 < z < 2.9) quasars. The data set includes four sources with integrated steep radio spectra and seven with flat radio spectra. A total of 25 radio jet features are present in this sample. We apply a Bayesian multi-scale image reconstruction method to detect and measure the X-ray emission from the jets. We compute deviations from a baseline model that does not include the jet, and compare observed X-ray images with those computed with simulated images where no jet features exist. This allows us to compute p -value upper bounds on the significance that an X-ray jet is detected in a pre-determined region of interest. We detected 12 of the features unambiguously, and an additional six marginally. We also find residual emission in the cores of three quasars and in the background of one quasar that suggest the existence of unresolved X-ray jets. The dependence of the X-ray to radio luminosity ratio on redshift is a potential diagnostic of the emission mechanism, since the inverse Compton scattering of cosmic microwave background photons (IC/CMB) is thought to be redshift dependent, whereas in synchrotron models no clear redshift dependence is expected. We find that the high-redshift jets have X-ray to radio flux ratios that are marginally inconsistent with those from lower redshifts, suggesting that either the X-ray emissions are due to the IC/CMB rather than the synchrotron process, or that high-redshift jets are qualitatively different.

79 ASTRONOMY AND ASTROPHYSICS↗

Hydrogeological Model Selection Among Complex Spatial Priors

Hydrogeological field studies rely often on a single conceptual representation of the subsurface. This is problematic since the impact of a poorly chosen conceptual model on predictions might be significantly larger than the one caused by parameter uncertainty. Furthermore, conceptual models often need to incorporate geological concepts and patterns in order to provide meaningful uncertainty quantification and predictions. Consequently, several geologically realistic conceptual models should ideally be considered and evaluated in terms of their relative merits. Here, we propose a full Bayesian methodology based on Markov chain Monte Carlo to enable model selection among 2-D conceptual models that are sampled using training images and concepts from multiple-point statistics. More precisely, power posteriors for the different conceptual subsurface models are sampled using sequential geostatistical resampling and Graph Cuts. To demonstrate the methodology, we compare and rank five alternative conceptual geological models that have been proposed in the literature to describe aquifer heterogeneity at the MAcroDispersion Experiment site in Mississippi, USA. We consider a small-scale tracer test for which the spatial distribution of hydraulic conductivity impacts multilevel solute concentration data observed along a 2-D transect. The thermodynamic integration and the stepping-stone sampling methods were used to compute the evidence and associated Bayes factors using the computed power posteriors. We find that both methods are compatible with multiple-point statistics-based inversions and provide a consistent ranking of the competing conceptual models considered.

58 GEOSCIENCES↗

Adaptive hyperparameter updating for training restricted Boltzmann machines on quantum annealers

Restricted Boltzmann Machines (RBMs) have been proposed for developing neural networks for a variety of unsupervised machine learning applications such as image recognition, drug discovery, and materials design. The Boltzmann probability distribution is used as a model to identify network parameters by optimizing the likelihood of predicting an output given hidden states trained on available data. Training such networks often requires sampling over a large probability space that must be approximated during gradient based optimization. Quantum annealing has been proposed as a means to search this space more efficiently which has been experimentally investigated on D-Wave hardware. D-Wave implementation requires selection of an effective inverse temperature or hyperparameter (β) within the Boltzmann distribution which can strongly influence optimization. Here, we show how this parameter can be estimated as a hyperparameter applied to D-Wave hardware during neural network training by maximizing the likelihood or minimizing the Shannon entropy. We find both methods improve training RBMs based upon D-Wave hardware experimental validation on an image recognition problem. Neural network image reconstruction errors are evaluated using Bayesian uncertainty analysis which illustrate more than an order magnitude lower image reconstruction error using the maximum likelihood over manually optimizing the hyperparameter. The maximum likelihood method is also shown to out-perform minimizing the Shannon entropy for image reconstruction.

97 MATHEMATICS AND COMPUTING↗

sMF-BO-2CoGP: A sequential multi-fidelity constrained Bayesian optimization framework for design applications

Bayesian optimization (BO) is an effective surrogate-based method that has been widely used to optimize simulation-based applications. While the traditional Bayesian optimization approach only applies to single-fidelity models, many realistic applications provide multiple levels of fidelity with various levels of computational complexity and predictive capability. In this work, we propose a multi-fidelity Bayesian optimization method for design applications with both known and unknown constraints. The proposed framework, called sMF-BO-2CoGP, is built on a multi-level CoKriging method to predict the objective function. An external binary classifier, which we approximate using a separate CoKriging model, is used to distinguish between feasible and infeasible regions. Finally, the sMF-BO-2CoGP method is demonstrated using a series of analytical examples and a flip-chip application for design optimization to minimize the deformation due to warping under thermal loading conditions.

42 ENGINEERING↗

Differentiable Multiphysics Codes: A Breakthrough Technology for Simulation and Computing

This document summarizes the findings of a strategic planning exercise commissioned by the Weapons Simulation and Computing, Computational Physics (WSC/CP) program at the Lawrence Livermore National Laboratory (LLNL) in FY24. During the year, the committee met with multiple stakeholder communities to gather input, opinions, suggestions and concerns which have been incorporated throughout this document. The key findings from this exercise are summarized: • The development of multiphysics modelling and simulation (mod/sim) codes and software technologies, their deployment on exascale compute platforms, and their broad adoption across the NNSA is a major success of the Advanced Simulation and Computing (ASC) program and the Exascale Computing Project (ECP). Sustained investment in these core technologies is essential. • Today’s state of the art involves running ensembles of O(100K) simulations to perform uncertainty quantification (UQ) and design studies using multiple statistical methods such as Bayesian optimization to understand sensitivities of our models and explore parameterized design spaces. Even with exascale computing, we are practically limited to O(10) parameters in these studies since the number of simulations required to sample the space scales exponentially with the number of design parameters. • The data from these simulation ensembles is increasingly being used to train machine learned (ML) surrogates (or reduced order models, ROMs) which can then be used for optimization or real time design exploration. However, the trained surrogates are still limited in the number of parameters they can represent due to the sampling limitations previously noted. • Augmenting our suite of integrated multiphysics simulation codes, both current and emerging, with the ability to compute gradients (solution derivatives) of arbitrary simulation outputs with respect to (some or all) simulation inputs would be a breakthrough technology, opening the door to a new era of efficient and automated inverse design based on verified and validated mod/sim capabilities. • This capability, which we refer to as differentiable multiphysics codes (DMCs), would revolutionize both UQ and optimization studies by breaking the curse of dimensionality that presently limits our “gradient-free” ensemble based computing approach. A similar breakthrough occurred in the AI/ML community once the ability to compute gradients of arbitrary loss functions using back-propagation became commonplace. Gradient information from the multiphysics codes can also be used to dramatically improve the efficiency and scale of training of ML/ROM surrogates for rapid assessments. • Achieving this in our suite of codes will be a grand challenge, similar to the amount of effort that was required to transition from CPU to GPU computing. It will require buy-in from the entire WSC/CP program and beyond, including all integrated codes, physics and engineering models, third-party library dependencies and performance portability abstractions. It will also require investment in research and development of numerical methods for computing adjoints of coupled physics across multiple adaptively refined moving meshes and of stochastic (Monte Carlo) and mesh free (SPH) methods. • New software and numerical techniques, largely pioneered by the AI/ML community, make this feasible. Chief among these is automatic differentiation (AD), the ability to employ AD at point-wise locations in a physics calculation (instead of traditional black-box approaches) and the ability to perform “back-propagation in time” (or reverse mode AD) for non-linear partial differential equations (PDEs). Fundamentally, the conclusion of this strategic planning exercise is that the time is right to undertake a large scale effort in WSC, centered on the existing integrated codes, to continue the natural evolution of mod/sim in the age of AI/ML. Instead of attempting to replace mod/sim with purely data driven AI/ML models, we believe the key to success is to integrate AI/ML by building on top of the decades of hard-won knowledge and the verified/validated multiphysics modelling capability that is the hallmark of the ASC program.

97 MATHEMATICS AND COMPUTING↗

Bayesian Model Calibration for Extrapolative Prediction via Gibbs Posteriors

The current standard Bayesian approach to model calibration, which assigns a Gaussian process prior to the discrepancy term, often suffers from issues of unidentifiability and computational complexity and instability. When the goal is to quantify uncertainty in physical parameters for extrapolative prediction, then there is no need to perform inference on the discrepancy term. With this in mind, we introduce Gibbs posteriors as an alternative Bayesian method for model calibration, which updates the prior with a loss function connecting the data to the parameter. The target of inference is the physical parameter value which minimizes the expected loss. We propose to tune the loss scale of the Gibbs posterior to maintain nominal frequentist coverage under assumptions of the form of model discrepancy, and present a bootstrap implementation for approximating coverage rates. Our approach is highly modular, allowing an analyst to easily encode a wide variety of such assumptions. Furthermore, we provide a principled method of combining posteriors calculated from data subsets. We apply our methods to data from an experiment measuring the material properties of tantalum.

36 MATERIALS SCIENCE↗

Search for very-short-baseline oscillations of reactor antineutrinos with the SoLid detector

In this paper we report the first scientific result based on antineutrinos emitted from the BR2 reactor at SCK CEN. The SoLid experiment uses a novel type of highly granular detector whose basic detection unit combines two scintillators, polyvinyl toluene (PVT) and Li 6 F : ZnS ( Ag ) , to measure antineutrinos via their inverse- β -decay products. An advantage of PVT is its highly linear response as a function of deposited particle energy. The full-scale detector comprises 12 800 voxels and operates over a very short 6.3–8.9 m baseline from the reactor core. The detector segmentation and its three-dimensional imaging capabilities facilitate the extraction of the positron energy from the rest of the visible energy, allowing the latter to be utilized for signal-background discrimination. We present a result obtained from 280 reactor-on days (55 MW mean power) and 172 reactor-off days, respectively, of live data taking. A total of 29 479 ± 603 (stat) antineutrino candidates have been selected, corresponding to an average rate of 105 events per day and a signal-to-background ratio of 0.27. A search for disappearance of antineutrinos to a sterile state has been conducted using complementary model-dependent frequentist and Bayesian fits, providing constraints on the allowed region of the reactor antineutrino anomaly.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

ExoReL{sup R}: A Bayesian Inverse Retrieval Framework for Exoplanetary Reflected Light Spectra

The high-contrast imaging technique is meant to provide insight into those planets orbiting several astronomical units from their host star. Space missions such as Wide-Field InfraRed Survey Telescope, Habitable Exoplanet Imaging Mission, and Large Ultra-Violet/Optical/InfraRed Surveyor will measure reflected light spectra of cold gaseous and rocky planets. To interpret these observations, we introduce ExoReL{sup R} (Exoplanetary Reflected Light Retrieval), a novel Bayesian retrieval framework to retrieve cloud properties and atmospheric structures from exoplanetary reflected light spectra. As a unique feature, it assumes a vertically nonuniform volume mixing ratio (VMR) profile of water and ammonia, and uses it to construct cloud densities. In this way, clouds and molecular mixture ratios are consistent. We apply ExoReL{sup R} on three test cases: two exoplanets (υ And e and 47 Uma b) and Jupiter. We show that we are able to retrieve the concentration of methane in the atmosphere, and estimate the position of clouds when the signal-to-noise ratio of the spectrum is higher than 15, in line with previous works. Moreover, we described the ability of our model to give a chemical identity to clouds, and we discussed whether or not we can observe this difference in the planetary reflection spectrum. Finally, we demonstrate how it could be possible to retrieve molecular concentrations (water and ammonia in this work) below the clouds by linking the nonuniform VMR profile to the cloud presence. This will help to constrain the concentration of water and ammonia unseen in direct measurements.

79 ASTRONOMY AND ASTROPHYSICS↗

BAYESIAN MAGNETOHYDRODYNAMIC SEISMOLOGY OF CORONAL LOOPS

We perform a Bayesian parameter inference in the context of resonantly damped transverse coronal loop oscillations. The forward problem is solved in terms of parametric results for kink waves in one-dimensional flux tubes in the thin tube and thin boundary approximations. For the inverse problem, we adopt a Bayesian approach to infer the most probable values of the relevant parameters, for given observed periods and damping times, and to extract their confidence levels. The posterior probability distribution functions are obtained by means of Markov Chain Monte Carlo simulations, incorporating observed uncertainties in a consistent manner. We find well-localized solutions in the posterior probability distribution functions for two of the three parameters of interest, namely the Alfven travel time and the transverse inhomogeneity length scale. The obtained estimates for the Alfven travel time are consistent with previous inversion results, but the method enables us to additionally constrain the transverse inhomogeneity length scale and to estimate real error bars for each parameter. When observational estimates for the density contrast are used, the method enables us to fully constrain the three parameters of interest. These results can serve to improve our current estimates of unknown physical parameters in coronal loops and to test the assumed theoretical model.

79 ASTRONOMY AND ASTROPHYSICS↗

Risk Analysis Virtual ENvironment

RAVEN has 3 major functionalities: 1. Provides a Graphical User Interface for the pre- and post-processing of the RELAP-7 input and output. 2. Provides the capability to model nuclear power plants control logic for the RELAP-7 code and dynamic control of the accident scenario evolution. This capability is based on a software structure that realizes a direct connection between the RELAP-7 solver engine (MOOSE) and a python environment where the variables describing the plant status are accessible in a scripting environment. RAVEN support the generation of the probabilistic scenario control by supplying a wide range of probability and cumulative distribution functions and their inverse functions. 3. Provides a general environment to perform probability risk analysis for RELAP-7, RELAP-5 and any generic MOOSE based applications. The probabilistic analysis is performed by sampling the input space of the coupled code parameters and it is enhanced by using modern artificial intelligence algorithms that accelerate the identification of the areas of major risk (in the input parameter space). This environment also provides a graphical visualization capability to analyze the outcomes. Among other approaches, the classical Monte Carlo and Latin Hypercube sampling algorithms are available. For the acceleration of the convergence of the sampling methodologies, Support Vector Machines, Bayesian regression, and collocation stochastic polynomials chaos are implemented. The same methodologies here described could be used to solve optimization and uncertainties propagation problems using the RAVEN framework.

Rabiti, Cristian↗

Statistical Uncertainty in Paleoclimate Proxy Reconstructions

A quantitative analysis of any environment older than the instrumental record relies on proxies. Uncertainties associated with proxy reconstructions are often underestimated, which can lead to artificial conflict between different proxies, and between data and models. In this paper, using ordinary least squares linear regression as a common example, we describe a simple, robust and generalizable method for quantifying uncertainty in proxy reconstructions. We highlight the primary controls on the magnitude of uncertainty, and compare this simple estimate to equivalent estimates from Bayesian, nonparametric and fiducial statistical frameworks. We discuss when it may be possible to reduce uncertainties, and conclude that the unexplained variance in the calibration must always feature in the uncertainty in the reconstruction. This directs future research toward explaining as much of the variance in the calibration data as possible. We also advocate for a “data-forward” approach, that clearly decouples the presentation of proxy data from plausible environmental inferences.

58 GEOSCIENCES↗

Studies of regional-scale climate variability and change. Hidden Markov models and coupled ocean-atmosphere modes

This project was a continuation of previous work under DOE CCPP funding, in which we had developed a twin approach of probabilistic network (PN) models (sometimes called dynamic Bayesian networks) and intermediate-complexity coupled ocean-atmosphere models (ICMs) to identify the predictable modes of climate variability and to investigate their impacts on the regional scale. We had developed a family of PNs (similar to Hidden Markov Models) to simulate historical records of daily rainfall, and used them to downscale GCM seasonal predictions. Using an idealized atmospheric model, we had established a novel mechanism through which ocean-induced sea-surface temperature (SST) anomalies might influence large-scale atmospheric circulation patterns on interannual and longer time scales; we had found similar patterns in a hybrid coupled ocean-atmosphere-sea-ice model. The goal of the this continuation project was to build on these ICM results and PN model development to address prediction of rainfall and temperature statistics at the local scale, associated with global climate variability and change, and to investigate the impact of the latter on coupled ocean-atmosphere modes. Our main results from the grant consist of extensive further development of the hidden Markov models for rainfall simulation and downscaling together with the development of associated software; new intermediate coupled models; a new methodology of inverse modeling for linking ICMs with observations and GCM results; and, observational studies of decadal and multi-decadal natural climate results, informed by ICM results.

54 ENVIRONMENTAL SCIENCES↗

A stochastic approach to quantifying the blur with uncertainty estimation for high-energy X-ray imaging systems

One of the primary causes of blur in a high-energy X-ray imaging system is the shape and extent of the radiation source, or ‘spot’. It is important to be able to quantify the size of the spot as it provides a lower bound on the recoverable resolution for a radiograph, and penumbral imaging methods – which involve the analysis of blur caused by a structured aperture – can be used to obtain the spot’s spatial profile. We present a Bayesian approach for estimating the spot shape that, unlike variational methods, is robust to the initial choice of parameters. The posterior is obtained from a normal likelihood, which was constructed from a weighted least squares approximation to a Poisson noise model, and prior assumptions that enforce both smoothness and non-negativity constraints. A Markov chain Monte Carlo algorithm is used to obtain samples from the target posterior, and the reconstruction and uncertainty estimates are the computed mean and variance of the samples, respectively. Lastly, synthetic data-sets are used to demonstrate accurate reconstruction, while real data taken with high-energy X-ray imaging systems are used to demonstrate applicability and feasibility.

97 MATHEMATICS AND COMPUTING↗