Search NASA⌕ Search

SEARCH · Search NASA

Results for “Bayesian method”

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 595 records · Page 33

Simplifying Probability Elicitation and Uncertainty Modeling in Bayesian Networks

In this paper we contribute two methods that simplify the demands of knowledge elicitation for particular types of Bayesian networks. The first method simplify the task of providing probabilities when the states that a random variable takes can be described by a new, fully ordered state set in which a state implies all the preceding states. The second method leverages Dempster-Shafer theory of evidence to provide a way for the expert to express the degree of ignorance that they feel about the estimates being provided.

99 GENERAL AND MISCELLANEOUS↗

EVOLUTION OF THE VELOCITY-DISPERSION FUNCTION OF LUMINOUS RED GALAXIES: A HIERARCHICAL BAYESIAN MEASUREMENT

We present a hierarchical Bayesian determination of the velocity-dispersion function of approximately 430,000 massive luminous red galaxies observed at relatively low spectroscopic signal-to-noise ratio (S/N {approx} 3-5 per 69 km s{sup -1}) by the Baryon Oscillation Spectroscopic Survey (BOSS) of the Sloan Digital Sky Survey III. We marginalize over spectroscopic redshift errors, and use the full velocity-dispersion likelihood function for each galaxy to make a self-consistent determination of the velocity-dispersion distribution parameters as a function of absolute magnitude and redshift, correcting as well for the effects of broadband magnitude errors on our binning. Parameterizing the distribution at each point in the luminosity-redshift plane with a log-normal form, we detect significant evolution in the width of the distribution toward higher intrinsic scatter at higher redshifts. Using a subset of deep re-observations of BOSS galaxies, we demonstrate that our distribution-parameter estimates are unbiased regardless of spectroscopic S/N. We also show through simulation that our method introduces no systematic parameter bias with redshift. We highlight the advantage of the hierarchical Bayesian method over frequentist 'stacking' of spectra, and illustrate how our measured distribution parameters can be adopted as informative priors for velocity-dispersion measurements from individual noisy spectra.

79 ASTRONOMY AND ASTROPHYSICS↗

Challenges in Markov Chain Monte Carlo for Bayesian Neural Networks

Markov chain Monte Carlo (MCMC) methods have not been broadly adopted in Bayesian neural networks (BNNs). This paper initially reviews the main challenges in sampling from the parameter posterior of a neural network via MCMC. Such challenges culminate to lack of convergence to the parameter posterior. Nevertheless, this paper shows that a nonconverged Markov chain, generated via MCMC sampling from the parameter space of a neural network, can yield via Bayesian marginalization a valuable posterior predictive distribution of the output of the neural network. Further, classification examples based on multilayer perceptrons showcase highly accurate posterior predictive distributions. The postulate of limited scope for MCMC developments in BNNs is partially valid; an asymptotically exact parameter posterior seems less plausible, yet an accurate posterior predictive distribution is a tenable research avenue.

97 MATHEMATICS AND COMPUTING↗

IZI: INFERRING THE GAS PHASE METALLICITY (Z) AND IONIZATION PARAMETER (q) OF IONIZED NEBULAE USING BAYESIAN STATISTICS

We present a new method for inferring the metallicity (Z) and ionization parameter (q) of H II regions and star-forming galaxies using strong nebular emission lines (SELs). We use Bayesian inference to derive the joint and marginalized posterior probability density functions for Z and q given a set of observed line fluxes and an input photoionization model. Our approach allows the use of arbitrary sets of SELs and the inclusion of flux upper limits. The method provides a self-consistent way of determining the physical conditions of ionized nebulae that is not tied to the arbitrary choice of a particular SEL diagnostic and uses all the available information. Unlike theoretically calibrated SEL diagnostics, the method is flexible and not tied to a particular photoionization model. We describe our algorithm, validate it against other methods, and present a tool that implements it called IZI. Using a sample of nearby extragalactic H II regions, we assess the performance of commonly used SEL abundance diagnostics. We also use a sample of 22 local H II regions having both direct and recombination line (RL) oxygen abundance measurements in the literature to study discrepancies in the abundance scale between different methods. We find that oxygen abundances derived through Bayesian inference using currently available photoionization models in the literature can be in good (∼30%) agreement with RL abundances, although some models perform significantly better than others. We also confirm that abundances measured using the direct method are typically ∼0.2 dex lower than both RL and photoionization-model-based abundances.

79 ASTRONOMY AND ASTROPHYSICS↗

A comprehensive review of dwell time optimization methods in computer-controlled optical surfacing

Dwell time plays a vital role in determining the accuracy and convergence of the computer-controlled optical surfacing process. However, optimizing dwell time presents a challenge due to its ill-posed nature, resulting in non-unique solutions. To address this issue, several well-known methods have emerged, including the iterative, Bayesian, Fourier transform, and matrix-form methods. Despite their independent development, these methods share common objectives, such as minimizing residual errors, ensuring dwell time's positivity and smoothness, minimizing total processing time, and enabling flexible dwell positions. This paper aims to comprehensively review the existing dwell time optimization methods, explore their interrelationships, provide insights for their effective implementations, evaluate their performances, and ultimately propose a unified dwell time optimization methodology.

36 MATERIALS SCIENCE↗

Parameter Calibration for Johnson Cook and Preston-Tonks-Wallace Material Strength Models with Uncertainty Quantification

In this study we perform a Bayesian calibration of the parameters in the Johnson Cook (JC) material strength model, with uncertainty, using experiments with for a range of low and medium strain rates. For the parameter calibration, we used a variational Bayesian approach with experimental data from Hopkinson bar and quasi-static tests done on Oxygen Free High Conductivity (OFHC) copper. The estimated parameter values matched well with experimental data for a modest range of strain rates and temperatures. Through this method, we also recovered uncertainty and correlation information for the estimated parameter values. We also compared our results to a calibration of parameters in the Preston-Tonks-Wallace (PTW) material strength model, using the same variational Bayesian method. The parameters estimated for both models provided good agreement with the experimental data.

36 MATERIALS SCIENCE↗

Uncertainty quantification in breakup reactions

Breakup reactions are one of the favored probes to study loosely bound nuclei, particularly those in the limit of stability forming a halo. In order to interpret such breakup experiments, the continuum discretized coupled channel method is typically used. In this study, the first Bayesian analysis of a breakup reaction model is performed. Herein we use a combination of statistical methods together with a three-body reaction model (the continuum discretized coupled channel method) to quantify the uncertainties on the breakup observables due to the parameters in the effective potential describing the loosely bound projectile of interest. The combination of tools we develop opens the path for a Bayesian analysis of not only breakup processes, but also a wide array of complex processes that require computationally intensive reaction models.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Unfolding quantum computer readout noise

Abstract In the current era of noisy intermediate-scale quantum computers, noisy qubits can result in biased results for early quantum algorithm applications. This is a significant challenge for interpreting results from quantum computer simulations for quantum chemistry, nuclear physics, high energy physics (HEP), and other emerging scientific applications. An important class of qubit errors are readout errors. The most basic method to correct readout errors is matrix inversion, using a response matrix built from simple operations to probe the rate of transitions from known initial quantum states to readout outcomes. One challenge with inverting matrices with large off-diagonal components is that the results are sensitive to statistical fluctuations. This challenge is familiar to HEP, where prior-independent regularized matrix inversion techniques (“unfolding”) have been developed for years to correct for acceptance and detector effects, when performing differential cross section measurements. We study one such method, known as iterative Bayesian unfolding, as a potential tool for correcting readout errors from universal gate-based quantum computers. This method is shown to avoid pathologies from commonly used matrix inversion and least squares methods.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A COMPLETE SPECTROSCOPIC SURVEY OF THE MILKY WAY SATELLITE SEGUE 1: DARK MATTER CONTENT, STELLAR MEMBERSHIP, AND BINARY PROPERTIES FROM A BAYESIAN ANALYSIS

We introduce a comprehensive analysis of multi-epoch stellar line-of-sight velocities to determine the intrinsic velocity dispersion of the ultrafaint satellites of the Milky Way. Our method includes a simultaneous Bayesian analysis of both membership probabilities and the contribution of binary orbital motion to the observed velocity dispersion within a 14-parameter likelihood. We apply our method to the Segue 1 dwarf galaxy and conclude that Segue 1 is a dark-matter-dominated galaxy at high probability with an intrinsic velocity dispersion of 3.7{sup +1.4}{sub -1.1} km s{sup -1}. The dark matter halo required to produce this dispersion must have an average density of {rho}-bar{sub 1/2} = 2.5{sub -1.9}{sup +4.1}, M{sub sun}, pc{sup -3} within a sphere that encloses half the galaxy's stellar luminosity. This is the highest measured density of dark matter in the Local Group. Our results show that a significant fraction of the stars in Segue 1 may be binaries with the most probable mean period close to 10 years, but also consistent with the 180 year mean period seen in the solar vicinity at about 1{sigma}. Despite this binary population, the possibility that Segue 1 is a bound star cluster with the observed velocity dispersion arising from the orbital motion of binary stars is disfavored by the multi-epoch stellar velocity data at greater than 99% C.L. Finally, our treatment yields a projected (two-dimensional) half-light radius for the stellar profile of Segue 1 of R{sub 1/2} = 28{sup +5}{sub -4} pc, in excellent agreement with photometric measurements.

79 ASTRONOMY AND ASTROPHYSICS↗

QUASI-PERIODIC OSCILLATIONS AND BROADBAND VARIABILITY IN SHORT MAGNETAR BURSTS

The discovery of quasi-periodic oscillations (QPOs) in magnetar giant flares has opened up prospects for neutron star asteroseismology. However, with only three giant flares ever recorded, and only two with data of sufficient quality to search for QPOs, such analysis is seriously data limited. We set out a procedure for doing QPO searches in the far more numerous, short, less energetic magnetar bursts. The short, transient nature of these bursts requires the implementation of sophisticated statistical techniques to make reliable inferences. Using Bayesian statistics, we model the periodogram as a combination of red noise at low frequencies and white noise at high frequencies, which we show is a conservative approach to the problem. We use empirical models to make inferences about the potential signature of periodic and QPOs at these frequencies. We compare our method with previously used techniques and find that although it is on the whole more conservative, it is also more reliable in ruling out false positives. We illustrate our Bayesian method by applying it to a sample of 27 bursts from the magnetar SGR J0501+4516 observed by the Fermi Gamma-ray Burst Monitor, and we find no evidence for the presence of QPOs in any of the bursts in the unbinned spectra, but do find a candidate detection in the binned spectra of one burst. However, whether this signal is due to a genuine quasi-periodic process, or can be attributed to unmodeled effects in the noise is at this point a matter of interpretation.

79 ASTRONOMY AND ASTROPHYSICS↗

Calibration of a distributed flood forecasting model with input uncertainty using a Bayesian framework

In the process of calibrating distributed hydrological models, accounting for input uncertainty is important, yet challenging. In this study, we develop a Bayesian model to estimate parameters associated with a geomorphology‐based hydrological model (GBHM). The GBHM model uses geomorphic characteristics to simplify model structure and physically based methods to represent hydrological processes. We divide the observed discharge into low‐ and high‐flow data, and use the first‐order autoregressive model to describe their temporal dependence. We consider relative errors in rainfall as spatially distributed variables and estimate them jointly with the GBHM parameters. The joint posterior probability distribution is explored using Markov chain Monte Carlo methods, which include Metropolis‐Hastings, delay rejection adaptive Metropolis, and Gibbs sampling methods. We evaluate the Bayesian model using both synthetic and field data sets. The synthetic case study demonstrates that the developed method generally is effective in calibrating GBHM parameters and in estimating their associated uncertainty. The calibration ignoring input errors has lower accuracy and lower reliability compared to the calibration that includes estimation of the input errors, especially under model structure uncertainty. The field case study shows that calibration of GBHM parameters under complex field conditions remains a challenge. Although jointly estimating input errors and GBHM parameters improves the continuous ranked probability score and the consistency of the predictive distribution with the observed data, the improvement is incremental. To better calibrate parameters in a distributed model, such as GBHM here, we need to develop a more complex model and incorporate much more information.

54 ENVIRONMENTAL SCIENCES↗

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↗

Cosmology and the neutrino mass ordering

We propose a simple method to quantify a possible exclusion of the inverted neutrino mass ordering from cosmological bounds on the sum of the neutrino masses. The method is based on Bayesian inference and allows for a calculation of the posterior odds of normal versus inverted ordering. We apply the method for a specific set of current data from Planck CMB data and large-scale structure surveys, providing an upper bound on the sum of neutrino masses of 0.14 eV at 95% CL. With this analysis we obtain posterior odds for normal versus inverted ordering of about 2:1. If cosmological data is combined with data from oscillation experiments the odds reduce to about 3:2. For an exclusion of the inverted ordering from cosmology at more than 95% CL, an accuracy of better than 0.02 eV is needed for the sum. We demonstrate that such a value could be reached with planned observations of large scale structure by analysing artificial mock data for a EUCLID-like survey.

79 ASTRONOMY AND ASTROPHYSICS↗

PYOED: AN ETENSIBLE SUITE FOR DATA ASSIMILATION AND MODEL-CONSTRAINED OPTIMAL DESIGN OF EXPERIMENTS

SF-23-005 PyOED is a highly extensible scientific package that enables developing and testing model-constrained optimal experimental design (OED) for inverse problems. Specifically, PyOED aims to be a comprehensive Python toolkit for model-constrained OED. The package targets scientists and researchers interested in understanding the details of OED formulations and approaches. It is also meant to enable researchers to experiment with standard and innovative OED technologies with a wide range of test problems (e.g., simulation models). OED, inverse problems (e.g., Bayesian inversion), and data assimilation (DA) are closely related research fields, and their formulations overlap significantly. Thus, PyOED is continuously being expanded with a plethora of Bayesian inversion, DA, and OED methods as well as new scientific simulation models, observation error models, and observation operators. These pieces are added such that they can be permuted to enable testing OED methods in various settings of varying complexities. The PyOED core is completely written in Python and utilizes the inherent object-oriented capabilities; however, PyOED is meant to be extensible rather than scalable. Specifically, PyOED is developed to ``enable rapid development and benchmarking of OED methods with minimal coding effort and to maximize code reutilization.'' PyOED will be continuously expanded with a plethora of Bayesian inversion, DA, and OED methods as well as new scientific simulation models, observation error models, and observation operators.

ATTIA, AHMEDMOHAMED↗

Life Prediction and Classification of Failure Modes in Solid State Luminaires Using Bayesian Probabilistic Models

A new method has been developed for assessment of the onset of degradation in solid state luminaires to classify failure mechanisms by using metrics beyond lumen degradation that are currently used for identification of failure. Luminous Flux output, Correlated Color Temperature Data on Philips LED Lamps has been gathered under 85°C/85%RH till lamp failure. The acquired data has been used in conjunction with Bayesian Probabilistic Models to identify luminaires with onset of degradation much prior to failure through identification of decision boundaries between lamps with accrued damage and lamps beyond the failure threshold in the feature space. In addition luminaires with different failure modes have been classified separately from healthy pristine luminaires. It is expected that, the new test technique will allow the development of failure distributions without testing till L70 life for the manifestation of failure.

Lall, Pradeep↗

Sequential Monte Carlo for Cut-Bayesian Posterior Computation

We propose a sequential Monte Carlo (SMC) method to efficiently and accurately compute cut-Bayesian posterior quantities of interest, variations of standard Bayesian approaches constructed primarily to account for model misspecification. We prove finite sample concentration bounds for estimators derived from the proposed method along with a linear tempering extension and apply these results to a realistic setting where a computer model is misspecified. We then illustrate the SMC method for inference in a modular chemical reactor example that includes submodels for reaction kinetics, turbulence, mass transfer, and diffusion. The samples obtained are commensurate with a direct-sampling approach that consists of running multiple Markov chains, with computational efficiency gains using the SMC method. Overall, the SMC method presented yields a novel, rigorous approach to computing with cut-Bayesian posterior distributions.

97 MATHEMATICS AND COMPUTING↗

Nuclear Forensics Analysis with Missing and Uncertain Data

We have applied a new imputation-based method for analyzing incomplete data, called Monte Carlo Bayesian Database Generation (MCBDG), to the Spent Fuel Isotopic Composition (SFCOMPO) database. About 60% of the entries are absent for SFCOMPO. The method estimates missing values of a property from a probability distribution created from the existing data for the property, and then generates multiple instances of the completed database for training a machine learning algorithm. Uncertainty in the data is represented by an empirical or an assumed error distribution. The method makes few assumptions about the underlying data, and compares favorably against results obtained by replacing missing information with constant values.

97 MATHEMATICS AND COMPUTING↗

An efficient Bayesian data-worth analysis using a multilevel Monte Carlo method

Improving the understanding of subsurface systems and thus reducing prediction uncertainty requires collection of data. As the collection of subsurface data is costly, it is important that the data collection scheme is cost-effective. Design of a cost-effective data collection scheme, i.e., data-worth analysis, requires quantifying model parameter, prediction, and both current and potential data uncertainties. Assessment of these uncertainties in large-scale stochastic subsurface hydrological model simulations using standard Monte Carlo (MC) sampling or surrogate modeling is extremely computationally intensive, sometimes even infeasible. In this paper, we propose an efficient Bayesian data-worth analysis using a multilevel Monte Carlo (MLMC) method. Compared to the standard MC that requires a significantly large number of high-fidelity model executions to achieve a prescribed accuracy in estimating expectations, the MLMC can substantially reduce computational costs using multifidelity approximations. Since the Bayesian data-worth analysis involves a great deal of expectation estimation, the cost saving of the MLMC in the assessment can be outstanding. While the proposed MLMC-based data-worth analysis is broadly applicable, we use it for a highly heterogeneous two-phase subsurface flow simulation to select an optimal candidate data set that gives the largest uncertainty reduction in predicting mass flow rates at four production wells. Finally, the choices made by the MLMC estimation are validated by the actual measurements of the potential data, and consistent with the standard MC estimation. But compared to the standard MC, the MLMC greatly reduces the computational costs.

97 MATHEMATICS AND COMPUTING↗