Search NASA⌕ Search

SEARCH · Search NASA

Results for “Bayesian Methods”

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 1,711 records · Page 95

Breaking Computational Barriers: Real-time Analysis and Optimization with Large-scale Nonlinear Models via Model Reduction

Model reduction for dynamical systems is a promising approach for reducing the computational cost of large-scale physics-based simulations to enable high-fidelity models to be used in many- query (e.g., Bayesian inference) and near-real-time (e.g., fast-turnaround simulation) contexts. While model reduction works well for specialized problems such as linear time-invariant systems, it is much more difficult to obtain accurate, stable, and efficient reduced-order models (ROMs) for systems with general nonlinearities. This report describes several advances that enable nonlinear reduced-order models (ROMs) to be deployed in a variety of time-critical settings. First, we present an error bound for the Gauss-Newton with Approximated Tensors (GNAT) nonlinear model reduction technique. This bound allows the state-space error for the GNAT method to be quantified when applied with the backward Euler time-integration scheme. Second, we present a methodology for preserving classical Lagrangian structure in nonlinear model reduction. This technique guarantees that important properties--such as energy conservation and symplectic time-evolution maps--are preserved when performing model reduction for models described by a Lagrangian formalism (e.g., molecular dynamics, structural dynamics). Third, we present a novel technique for decreasing the temporal complexity --defined as the number of Newton-like iterations performed over the course of the simulation--by exploiting time-domain data. Fourth, we describe a novel method for refining projection-based reduced-order models a posteriori using a goal-oriented framework similar to mesh-adaptive h -refinement in finite elements. The technique allows the ROM to generate arbitrarily accurate solutions, thereby providing the ROM with a 'failsafe' mechanism in the event of insufficient training data. Finally, we present the reduced-order model error surrogate (ROMES) method for statistically quantifying reduced- order-model errors. This enables ROMs to be rigorously incorporated in uncertainty-quantification settings, as the error model can be treated as a source of epistemic uncertainty. This work was completed as part of a Truman Fellowship appointment. We note that much additional work was performed as part of the Fellowship. One salient project is the development of the Trilinos-based model-reduction software module Razor , which is currently bundled with the Albany PDE code and currently allows nonlinear reduced-order models to be constructed for any application supported in Albany. Other important projects include the following: 1. ROMES-equipped ROMs for Bayesian inference: K. Carlberg, M. Drohmann, F. Lu (Lawrence Berkeley National Laboratory), M. Morzfeld (Lawrence Berkeley National Laboratory). 2. ROM-enabled Krylov-subspace recycling: K. Carlberg, V. Forstall (University of Maryland), P. Tsuji, R. Tuminaro. 3. A pseudo balanced POD method using only dual snapshots: K. Carlberg, M. Sarovar. 4. An analysis of discrete v. continuous optimality in nonlinear model reduction: K. Carlberg, M. Barone, H. Antil (George Mason University). Journal articles for these projects are in progress at the time of this writing.

97 MATHEMATICS AND COMPUTING↗

Decision trees for incorporating the likelihood of excessive hydrologic alteration into hydropower-ecosystem tradeoffs

Detrimental ecological impacts have often been observed or anticipated when changes in streamflow indicators exceed percent deviation thresholds believed to be ecologically critical. Yet, short pre- and post-impact flow records often make it difficult to determine whether changes exceeding tolerable thresholds are due to dam operations or natural variability. Through a hypothetical reservoir operations example, we incorporate the uncertainty of dam-induced streamflow changes into a Bayesian decision tree framework that evaluates tradeoffs between expected regrets associated with hydropower and ecology. The likelihood of over-protection (type I) and under-protection (type II) errors associated with hypothesis tests are used to compute expected hydropower and ecosystem regrets associated with dam operation decisions. We examine changes to high (annual Q5) and low flows (annual Q95) in typical years using a modified and extended nonparametric ranked-sum test that accounts for percent deviation thresholds. A multiple comparison test is then used to determine the likelihood of at least one threshold violation. An example shows that our decision-theoretic approach can lead to different dam operation recommendations than do other common methods, and highlights limitations that arise when the type I error rate is selected a priori. While we illustrate a hydropower-ecosystem tradeoff, our approach can also be applied to other multi-stakeholder reservoir and river management conflicts.

Bayesian decision-making, environmental flows, flo↗

Robust In-Situ Strain Measurements to Monitor CO 2 Storage

The goal of this project was to develop and demonstrate robust instrumentation to monitor the in-situ strain tensor in order to improve the reliability and security of CO 2 storage in geologic formations. We met the original goals of the project and the major overarching accomplishment is the advancement of strain tensor monitoring from an intriguing concept to a commercially available technology with a solid foundation of novel instruments supported by theoretical analyses and validation experiments. The main accomplishments of the project are summarized below. We designed, built and evaluated nine new optical fiber strainmeters and tiltmeters using Michelson interferometers to measure deformation with ultra-high resolution at both shallow and deep point locations in the subsurface. These are the robust strainmeters that motivated the title of the project. We designed, built and evaluated a novel method of measuring distributed strain in optical fibers with nanostrain resolution, and cm-scale location, and sampling into the seismic band. The new method is called Coherence-length-gated Microwave Photonics Interfereometry (CMPI). CMPI technology has advantages over existing commercial DAS and DSS methods. We developed and demonstrated capabilities to deploy instruments in the field and used them to measure strain caused by ambient signals like barometric pressure and tides, as well as induced signals like surface loading and pore pressure changes from pumping tests. We deployed a working strainmeter at 1,700 ft depth, slightly above an active reservoir. This is to our knowledge the greatest depth a strainmeter has been deployed and the techniques we used can readily be extended to greater depths. Optical fiber borehole tensor strainmeter techology was advanced from a TRL 4 at the start, to a TRL of 7 at the conclusion of the project. The project included advances in simulations and theoretical analyses. We developed and demonstrated a computational workflow that uses machine learning to reduce the computational requirements and make it practical to use Bayesian inversion to solve large numerical poroelastic analyses needed to interpret strain tensor field data. We evaluated the strain tensor fields and time series that would be caused by leaks of CO 2 or other fluids from reservoirs. These simulations demonstrated that signals from leaks could be measured with instruments developed for the project, opening a potentially new method for ensuring storage security. We showed that strains in caprock can be used to estimate pressure in a reservoir. This avoids the need to drill monitoring wells into the reservoir, and it expands the capabilities of monitoring in the caprock. The project includes a derivation and application of a novel analytical solution to the strains in the vicinity of a pressurized poroelastic inclusion. This solution explains field data measured during injeciton tests at the North Avant Field, and it will simplify future interpretation of strain tensor data. The project included a broad range of experiments, and of the most significant is the characterization of the strain tensor at an array three strainmeters during six injection tests at the North Avant Field, Oklahoma. This demonstrated repeatability of the strain signal measured by the new instruments developed for the project, and it showed similarities between the strain signal at shallow depths and pressure in the underlying reservoir. We also demonstrated that useful strain data can be measured at reservoir depths. This confirms that strain tensor data can be measured throughout the caprock over a reservoir. The project demonstrated the feasibility of using the strain tensor and distributed strain measured in caprock during a variety of different well tests where the pumping rate was constant, sinusoidal and positive, or a periodic square wave with zero net rate. This further strengthens the validity of using strain data to characterize reservoirs and aquifers. We also demonstrated that strain caused be fluctuations of air pressure and water pressure in the vadose zone can be measured and interpreted, suggesting that high resolution distributed strain measurements hold promise for monitoring the vadose zone. The project partially supported nine graduate students in the Environmental Engineering, Hydrogeology, Electrical Engineering programs at Clemson University. The research was described in nine journal papers, 23 talks and conference abstracts. Additional journal papers are in preparation. A new company called Tensora was started to provide strainmeter technology for commercial applications.

01 COAL, LIGNITE, AND PEAT↗

A Statistical Evaluation of Rules for Biochemical Failure After Radiotherapy in Men Treated for Prostate Cancer

Purpose: The 'PSA nadir + 2 rule,' defined as any rise of 2 ng/ml above the current prostate-specific antigen (PSA) nadir, has replaced the American Society for Therapeutic Radiology and Oncology (ASTRO) rule, defined as three consecutive PSA rises, to indicate biochemical failure (BF) after radiotherapy in patients treated for prostate cancer. We propose an original approach to evaluate BF rules based on the PSAdt as the gold standard rule and on a simulation process allowing us to evaluate the BF rules under multiple settings (different frequency, duration of follow-up, PSA doubling time [PSAdt]). Methods and Materials: We relied on a retrospective, population-based cohort of individuals identified by the Connecticut Tumor Registry and treated for localized prostate cancer with radiotherapy. We estimated the 470 underlying true PSA trajectories, including the PSAdt, using a Bayesian hierarchical changepoint model. Next, we simulated realistic, sophisticated data sets that accurately reflect the systematic and random variations observed in PSA series. We estimated the sensitivity and specificity by comparing the simulated PSA series to the underlying true PSAdt. Results: For follow-up of more than 3 years, the specificity of the PSA nadir + 2 rule was systematically greater than that of the ASTRO criterion. In few settings, the nadir + 2 rule had a lower sensitivity than the ASTRO. The PSA nadir + 2 rule appeared less dependent on the frequency and duration of follow-up than the ASTRO. Conclusions: Our results provide some refinements to earlier findings as the BF rules were evaluated according to various parameters. In most settings, the PSA nadir + 2 rule outperforms the ASTRO criterion.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Redshift inference from the combination of galaxy colours and clustering in a hierarchical Bayesian model – Application to realistic N -body simulations

ABSTRACT Photometric galaxy surveys constitute a powerful cosmological probe but rely on the accurate characterization of their redshift distributions using only broad-band imaging, and can be very sensitive to incomplete or biased priors used for redshift calibration. A hierarchical Bayesian model has recently been developed to estimate those from the robust combination of prior information, photometry of single galaxies, and the information contained in the galaxy clustering against a well-characterized tracer population. In this work, we extend the method so that it can be applied to real data, developing some necessary new extensions to it, especially in the treatment of galaxy clustering information, and we test it on realistic simulations. After marginalizing over the mapping between the clustering estimator and the actual density distribution of the sample galaxies, and using prior information from a small patch of the survey, we find the incorporation of clustering information with photo-z’s tightens the redshift posteriors and overcomes biases in the prior that mimic those happening in spectroscopic samples. The method presented here uses all the information at hand to reduce prior biases and incompleteness. Even in cases where we artificially bias the spectroscopic sample to induce a shift in mean redshift of $\Delta \bar{z} \approx 0.05,$ the final biases in the posterior are $\Delta \bar{z} \lesssim 0.003.$ This robustness to flaws in the redshift prior or training samples would constitute a milestone for the control of redshift systematic uncertainties in future weak lensing analyses.

79 ASTRONOMY AND ASTROPHYSICS↗

Testing general relativity using Bayesian model selection: Applications to observations of gravitational waves from compact binary systems

Second-generation interferometric gravitational-wave detectors, such as Advanced LIGO and Advanced Virgo, are expected to begin operation by 2015. Such instruments plan to reach sensitivities that will offer the unique possibility to test general relativity in the dynamical, strong-field regime and investigate departures from its predictions, in particular, using the signal from coalescing binary systems. We introduce a statistical framework based on Bayesian model selection in which the Bayes factor between two competing hypotheses measures which theory is favored by the data. Probability density functions of the model parameters are then used to quantify the inference on individual parameters. We also develop a method to combine the information coming from multiple independent observations of gravitational waves, and show how much stronger inference could be. As an introduction and illustration of this framework-and a practical numerical implementation through the Monte Carlo integration technique of nested sampling-we apply it to gravitational waves from the inspiral phase of coalescing binary systems as predicted by general relativity and a very simple alternative theory in which the graviton has a nonzero mass. This method can (and should) be extended to more realistic and physically motivated theories.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Application and Evaluation of Surrogate Models for Radiation Source Search

Surrogate models are increasingly required for applications in which first-principles simulation models are prohibitively expensive to employ for uncertainty analysis, design, or control. They can also be used to approximate models whose discontinuous derivatives preclude the use of gradient-based optimization or data assimilation algorithms. We consider the problem of inferring the 2D location and intensity of a radiation source in an urban environment using a ray-tracing model based on Boltzmann transport theory. Whereas the code implementing this model is relatively efficient, extension to 3D Monte Carlo transport simulations precludes subsequent Bayesian inference to infer source locations, which typically requires thousands to millions of simulations. Additionally, the resulting likelihood exhibits discontinuous derivatives due to the presence of buildings. To address these issues, we discuss the construction of surrogate models for optimization, Bayesian inference, and uncertainty propagation. Specifically, we consider surrogate models based on Legendre polynomials, multivariate adaptive regression splines, radial basis functions, Gaussian processes, and neural networks. We detail strategies for computing training points and discuss the merits and deficits of each method.

Cook, Jared A.↗

Hybrid Gibbs Sampling and MCMC for CMB Analysis at Small Angular Scales

A) Gibbs Sampling has now been validated as an efficient, statistically exact, and practically useful method for "low-L" (as demonstrated on WMAP temperature polarization data). B) We are extending Gibbs sampling to directly propagate uncertainties in both foreground and instrument models to total uncertainty in cosmological parameters for the entire range of angular scales relevant for Planck. C) Made possible by inclusion of foreground model parameters in Gibbs sampling and hybrid MCMC and Gibbs sampling for the low signal to noise (high-L) regime. D) Future items to be included in the Bayesian framework include: 1) Integration with Hybrid Likelihood (or posterior) code for cosmological parameters; 2) Include other uncertainties in instrumental systematics? (I.e. beam uncertainties, noise estimation, calibration errors, other).

Gibbs sampling↗

Data-Driven Strategies for Accelerated Materials Design

The ongoing revolution of the natural sciences by the advent of machine learning and artificial intelligence sparked significant interest in the material science community in recent years. The intrinsically high dimensionality of the space of realizable materials makes traditional approaches ineffective for large-scale explorations. Modern data science and machine learning tools developed for increasingly complicated problems are an attractive alternative. An imminent climate catastrophe calls for a clean energy transformation by overhauling current technologies within only several years of possible action available. Tackling this crisis requires the development of new materials at an unprecedented pace and scale. For example, organic photovoltaics have the potential to replace existing silicon-based materials to a large extent and open up new fields of application. In recent years, organic light-emitting diodes have emerged as state-of-the-art technology for digital screens and portable devices and are enabling new applications with flexible displays. Reticular frameworks allow the atom-precise synthesis of nanomaterials and promise to revolutionize the field by the potential to realize multifunctional nanoparticles with applications from gas storage, gas separation, and electrochemical energy storage to nanomedicine. In the recent decade, significant advances in all these fields have been facilitated by the comprehensive application of simulation and machine learning for property prediction, property optimization, and chemical space exploration enabled by considerable advances in computing power and algorithmic efficiency. In this Account, we review the most recent contributions of our group in this thriving field of machine learning for material science. We start with a summary of the most important material classes our group has been involved in, focusing on small molecules as organic electronic materials and crystalline materials. Specifically, we highlight the data-driven approaches we employed to speed up discovery and derive material design strategies. Subsequently, our focus lies on the data-driven methodologies our group has developed and employed, elaborating on high-throughput virtual screening, inverse molecular design, Bayesian optimization, and supervised learning. We discuss the general ideas, their working principles, and their use cases with examples of successful implementations in data-driven material discovery and design efforts. Furthermore, we elaborate on potential pitfalls and remaining challenges of these methods. Finally, we provide a brief outlook for the field as we foresee increasing adaptation and implementation of large scale data-driven approaches in material discovery and design campaigns.

36 MATERIALS SCIENCE↗

Bayesian tomographic reconstruction of microsystems

The microtomography by X ray transmission plays an increasingly dominating role in the study and the understanding of microsystems. Within this framework, an experimental setup of high resolution X ray microtomography was developed at CEA-List to quantify the physical parameters related to the fluids flow in microsystems. Several difficulties rise from the nature of experimental data collected on this setup: enhanced error measurements due to various physical phenomena occurring during the image formation (diffusion, beam hardening), and specificities of the setup (limited angle, partial view of the object, weak contrast).To reconstruct the object we must solve an inverse problem. This inverse problem is known to be ill-posed. It therefore needs to be regularized by introducing prior information. The main prior information we account for is that the object is composed of a finite known number of different materials distributed in compact regions. This a priori information is introduced via a Gauss-Markov field for the contrast distributions with a hidden Potts-Markov field for the class materials in the Bayesian estimation framework. The computations are done by using an appropriate Markov Chain Monte Carlo (MCMC) technique.In this paper, we present first the basic steps of the proposed algorithms. Then we focus on one of the main steps in any iterative reconstruction method which is the computation of forward and adjoint operators (projection and backprojection). A fast implementation of these two operators is crucial for the real application of the method. We give some details on the fast computation of these steps and show some preliminary results of simulations.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Cosmological limits on axions

We derive cosmological limits on two-component hot dark matter consisting of neutrinos and axions. We restrict the large-scale structure data to the safely linear regime, excluding the Lyman-{alpha} forest. We derive Bayesian credible regions in the two-parameter space consisting of m{sub a} and {sigma}m{sub v}. Marginalising over {sigma}m{sub v} provides m{sub a}<1.02 eV(95% C.L.). In the absence of axions the same data and methods give {sigma}m{sub v}<0.63 eV(95% C.L.)

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Almost medium-free measurement of the Hoyle state direct-decay component with a TPC

The structure of the Hoyle state, a highly α -clustered state at 7.65 MeV in 12 C, has long been the subject of debate. Understanding if the system comprises of three weakly interacting α particles in the 0s orbital, known as an α-condensate state, is possible by studying the decay branches of the Hoyle state. The direct decay of the Hoyle state into three α particles, rather than through the 8 Be ground state, can be identified by studying the energy partition of the three α particles arising from the decay. This paper provides details on the breakup mechanism of the Hoyle stating using a new experimental technique. Method: By using β-delayed charged-particle spectroscopy of 12 N using the Texas active target time-projection chamber, a high-sensitivity measurement of the direct 3α decay ratio can be performed without contributions from pileup events. A Bayesian approach to understanding the contribution of the direct components via a likelihood function shows that the direct component is <0.043% at the 95% confidence level. This value is in agreement with several other studies, and, here, we can demonstrate that a small nonsequential component with a decay fraction of about 10 –4 is most likely. Here, the measurement of the non-sequential component of the Hoyle state decay is performed in an almost medium-free reaction for the first time. The derived upper limit is in agreement with previous studies and demonstrates sensitivity to the absolute branching ratio. Further experimental studies would need to be combined with robust microscopic theoretical understanding of the decay dynamics to provide additional insight into the idea of the Hoyle state as an α condensate.

6 ≤ A ≤ 19↗

Real-time Event Detection Using Rank Signatures of Real-world PMU Data

Timely detection of power system events is a crucial task, which can facilitate the implementation of remedial actions to improve reliability, resiliency, and security of the system. Meanwhile, the widespread deployment of phasor measurement units (PMUs) makes it possible to develop data-driven event detection techniques. However, relying purely on data without incorporating domain knowledge for the event detection task in power systems poses substantial security and stability risks due to issues associated with data misinterpretation and model accuracy. In this regard, we propose a real-time event detection method using real-world PMU data by incorporating domain knowledge to adequately capture the event signatures. Specifically, we track the change in rank signatures of PMU data to accurately localize the events. To optimize the detection process, we incorporate an offline Bayesian optimization algorithm to tune the parameters by efficiently searching for the best values. The experiments using the real-world PMU dataset from a U.S. interconnection show that the proposed event detection approach can efficiently detect the events from PMU data streams with high accuracy.

Ghasemkhani, Amir↗

Understanding Material Strength Variabilities and Uncertainties for Component Qualification (2020 LDRD Final Report)

At its core, material specification and materials qualification require an understanding of the uncertainties in relation to a given performance metric. When simulations are used to inform the design and evaluation of these metrics, it is essential to understand the uncertainties in the underlying models and parameters that feed into those simulations. In this work, we have created a statistical framework to address uncertainties in the materials strength modeling used in integrated simulations. The framework is based on the Bayesian methodology, which allows for uncertainties to be updated as new data become available. The results of this work show how model parameters and their uncertainties can be updated as new and different types of strength data are added, as well as methods for determining which future experiment has the greatest potential to reduce the uncertainty. The statistical framework developed in this project will be utilized to assess and propagate the uncertainties in the strength models used in the programmatic simulations and to create a way to formally update those uncertainties as new data become available.

36 MATERIALS SCIENCE↗

Correlations between the Neutron Star Mass–Radius Relation and the Equation of State of Dense Matter

We develop an analytic method of inverting the Tolman–Oppenheimer–Volkoff relations to high accuracy. In principle, a specified energy density–pressure relation gives a unique mass–radius (M–R) relation and vice versa. Our method is developed from the strong correlations that are shown to exist between the neutron star mass–radius curve and the equation of state (EOS) or pressure–energy density relation. Selecting points that have masses equal to fixed fractions of the maximum mass, we find a semi-universal power-law relation between the central energy densities, pressures, sound speeds, chemical potentials, and number densities of those stars, with the maximum mass and the radii of one or more fractional maximum mass points. Rms fitting accuracies, for EOSs without large first-order phase transitions, are typically 0.5% for all quantities at all mass points. The method also works well, although less accurately, in reconstructing the EOS of hybrid stars with first-order phase transitions. These results permit, in effect, an analytic method of inverting an arbitrary M–R curve to yield its underlying EOS. We discuss applications of this inversion technique to the inference of the dense matter EOS from measurements of neutron star masses and radii as a possible alternative to traditional Bayesian approaches.

Bayesian statistics↗

Modeling the Swift BAT Trigger Algorithm with Machine Learning

To draw inferences about gamma-ray burst (GRB) source populations based on Swift observations, it is essential to understand the detection efficiency of the Swift burst alert telescope (BAT). This study considers the problem of modeling the Swift BAT triggering algorithm for long GRBs, a computationally expensive procedure, and models it using machine learning algorithms. A large sample of simulated GRBs from Lien et al. (2014) is used to train various models: random forests, boosted decision trees (with AdaBoost), support vector machines, and artificial neural networks. The best models have accuracies of approximately greater than 97% (approximately less than 3% error), which is a significant improvement on a cut in GRB flux which has an accuracy of 89:6% (10:4% error). These models are then used to measure the detection efficiency of Swift as a function of redshift z, which is used to perform Bayesian parameter estimation on the GRB rate distribution. We find a local GRB rate density of eta(sub 0) approximately 0.48(+0.41/-0.23) Gpc(exp -3) yr(exp -1) with power-law indices of eta(sub 1) approximately 1.7(+0.6/-0.5) and eta(sub 2) approximately -5.9(+5.7/-0.1) for GRBs above and below a break point of z(sub 1) approximately 6.8(+2.8/-3.2). This methodology is able to improve upon earlier studies by more accurately modeling Swift detection and using this for fully Bayesian model fitting. The code used in this is analysis is publicly available online.

gamma rays: general↗

A Singular Value Decomposition Framework for Retrievals with Vertical Distribution Information from Greenhouse Gas Column Absorption Spectroscopy Measurements

We review the singular value decomposition (SVD) framework and use it for quantifying and discerning vertical information in greenhouse gas retrievals from column integrated absorption measurements. While the commonly used traditional Bayesian optimal estimation (OE) assumes a prior distribution in order to regularize the inversion problem, the SVD approach identifies principal components that can be retrieved from the measurement without explicitly specifying a prior mean and prior covariance matrix. We review the SVD method, explicitly recognize the use of an uninformative prior and show it to incur no bias from the choice of the prior. We also make the connection between the SVD method and the pseudo-inverse, which makes it more intuitive and easy to understand. We illustrate the use of the SVD method on an integrated path differential absorption CO2 lidar measurement model and verify our derivations and bias-free properties versus optimal estimation using numerical simulations. In contrast, traditional OE retrievals exhibit bias when the prior mean used in the retrieval differs from the true mean. Hence, the SVD method is particularly useful for situations in which knowledge of the prior mean and prior covariance of the true state (e.g., greenhouse gas profiles) is inadequate.

Ramanathan, Anand K.↗