Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “probabilistic inference”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 181 records · Page 10

Phylogenetically resolving epidemiologic linkage

The use of phylogenetic trees in epidemiological investigations has become commonplace, but their epidemiological interpretation has not been systematically evaluated. Here, we use an HIV-1 within-host coalescent model to probabilistically evaluate transmission histories of two epidemiologically linked hosts. Previous critique of phylogenetic reconstruction has claimed that direction of transmission is difficult to infer, and that the existence of unsampled intermediary links or common sources can never be excluded. The phylogenetic relationship between the HIV populations of epidemiologically linked hosts can be classified into six types of trees, based on cladistic relationships and whether the reconstruction is consistent with the true transmission history or not. We show that the direction of transmission and whether unsampled intermediary links or common sources existed make very different predictions about expected phylogenetic relationships: (i) Direction of transmission can often be established when paraphyly exists, (ii) intermediary links can be excluded when multiple lineages were transmitted, and (iii) when the sampled individuals’ HIV populations both are monophyletic a common source was likely the origin. Inconsistent results, suggesting the wrong transmission direction, were generally rare. In addition, the expected tree topology also depends on the number of transmitted lineages, the sample size, the time of the sample relative to transmission, and how fast the diversity increases after infection. Typically, 20 or more sequences per subject give robust results. Moreover, we confirm our theoretical evaluations with analyses of real transmission histories and discuss how our findings should aid in interpreting phylogenetic results.

59 BASIC BIOLOGICAL SCIENCES↗

Imprecise global sensitivity analysis using bayesian multimodel inference and importance sampling

Global Sensitivity Analysis (GSA) aims to understand the relative importance of uncertain input variables to model response. Conventional GSA involves calculating sensitivity (Sobol’) indices for a model with known model parameter distributions. However, model parameters are affected by aleatory and epistemic uncertainty, with the latter often caused by lack of data. In this paper, we propose a new framework to quantify uncertainty in probability model-form and model parameters resulting from small datasets and integrate these uncertainties into Sobol’ index estimates. First, the process establishes, through Bayesian multimodel inference, a set of candidate probability models and their associated probabilities. Imprecise Sobol’ indices are calculated from these probability models using an importance sampling reweighting approach. This results in probabilistic Sobol’ indices, whose distribution characterizes uncertainty in the sensitivity resulting from small dataset size. The imprecise Sobol’ indices thus provide a measure of confidence in the sensitivity estimate and, moreover, can be used to inform data collection efforts targeted to minimize the impact of uncertainties. Through an example studying the parameters of a Timoshenko beam, we show that these probabilistic Sobol’ indices converge to the true/deterministic Sobol’ indices as the dataset size increases and hence, distribution-form uncertainty reduces. The approach is then applied to assess the sensitivity of the out-of-plane properties of an E-glass fiber composite material to its constituent properties. This second example illustrates the approach for an important class of materials with wide-ranging applications when data may be lacking for some input parameters.

42 ENGINEERING↗

Classification image analysis: estimation and statistical inference for two-alternative forced-choice experiments

We consider estimation and statistical hypothesis testing on classification images obtained from the two-alternative forced-choice experimental paradigm. We begin with a probabilistic model of task performance for simple forced-choice detection and discrimination tasks. Particular attention is paid to general linear filter models because these models lead to a direct interpretation of the classification image as an estimate of the filter weights. We then describe an estimation procedure for obtaining classification images from observer data. A number of statistical tests are presented for testing various hypotheses from classification images based on some more compact set of features derived from them. As an example of how the methods we describe can be used, we present a case study investigating detection of a Gaussian bump profile.

Non-NASA Center↗

Thoughts Without Content are Empty, Intuitions Without Concepts are Blind - Determinism and Contingency Revisited

Was the emergence of life a predictable outcome of chemical evolution on earth? Could evolution produce life very different from ours? These are one of the oldest questions in the field of the origin of life that not only have broad philosophical implications but also impact how we approach the problem from the methodological standpoint. Framing the issue in terms of the dichotomy between contingency and determinism is not a fortunate because these two terms in their conventional meaning are neither mutually exclusive nor jointly exhaustive. Determinism, represented in natural sciences by Newtonian physics, relies on the assumption that every event is causally determined by a chain of previous events. In the context of the origin of life it means that once the initial conditions on the early earth have been specified further evolution follows inevitably. Considering uncertainties about conditions on the prebiotic earth, many plausible sets of initial conditions can be defined, each followed by a separate deterministic trajectory. This conventional understanding of determinism does not admit contingency. Further, it has no implications for evaluating how many sets of initial conditions lead to the emergence of life. It appears that a better framing of the problem is as follows: given plausible sets of initial conditions on the early earth how probable and broadly spread are evolutionary trajectories that lead to life? Instead of undertaking an impossible task of specifying microscopic initial conditions for all components of the system one uses a reduced representation of this system and specify only a small set of essential macroscopic parameters, values (or ranges of values) of which can be identified, inferred or estimated from experiment, theory or historical record. The following evolutionary trajectories are still governed by laws of physics and chemistry but become probabilistic and "contingency" is admitted as variations in other variables in the system. A similar reasoning is common in other fields of science, for example in statistical mechanics. Some trajectories lead to life, perhaps in different forms, whereas others do not. Of our true interest is the ratio of these two outcomes. The issue of determinism does not directly enter the picture. The debate about the likelihood of the emergence of life is quite old. One view holds that the origin of life is an event governed by chance, and the result of so many random events (contingencies) is unpredictable. This view was eloquently expressed by Monod. In his book "Chance or Necessity" he argued that life was a product of "nature's roulette." In an alternative view, expressed in particular by deDuve and Morowitz, the origin of life is considered a highly probable or even inevitable event (although its details need not be determined in every respect). Only in this sense the origin of life can be considered a "deterministic event".

Pohorille, Andrew↗

Near Earth Asteroid Characterization for Threat Assessment

Physical characteristics of NEAs are an essential input to modeling behavior during atmospheric entry and to assess the risk of impact but determining these properties requires a non-trivial investment of time and resources. The characteristics relevant to these models include size, density, strength and ablation coefficient. Some of these characteristics cannot be directly measured, but rather must be inferred from related measurements of asteroids and/or meteorites. Furthermore, for the majority of NEAs, only the basic measurements exist so often properties must be inferred from statistics of the population of more completely characterized objects. The Asteroid Threat Assessment Project at NASA Ames Research Center has developed a probabilistic asteroid impact risk (PAIR) model in order to assess the risk of asteroid impact. Our PAIR model and its use to develop probability distributions of impact risk are discussed in other contributions to PDC 2017 (e.g., Mathias et al.). Here we utilize PAIR to investigate which NEA characteristics are important for assessing the impact threat by investigating how changes in these characteristics alter the damage predicted by PAIR. We will also provide an assessment of the current state of knowledge of the NEA characteristics of importance for asteroid threat assessment. The relative importance of different properties as identified using PAIR will be combined with our assessment of the current state of knowledge to identify potential high impact investigations. In addition, we will discuss an ongoing effort to collate the existing measurements of NEA properties of interest to the planetary defense community into a readily accessible database.

Asteroid characterization↗

Uncertainty quantification in the presence of limited climate model data with discontinuities.

Uncertainty quantification in climate models is challenged by the prohibitive cost of a large number of model evaluations for sampling. Another feature that often prevents classical uncertainty analysis from being readily applicable is the bifurcative behavior in the climate data with respect to certain parameters. A typical example is the Meridional Overturning Circulation in the Atlantic Ocean. The maximum overturning stream function exhibits a discontinuity across a curve in the space of two uncertain parameters, namely climate sensitivity and CO2 forcing. In order to propagate uncertainties from model parameters to model output we use polynomial chaos (PC) expansions to represent the maximum overturning stream function in terms of the uncertain climate sensitivity and CO2 forcing parameters. Since the spectral methodology assumes a certain degree of smoothness, the presence of discontinuities suggests that separate PC expansions on each side of the discontinuity will lead to more accurate descriptions of the climate model output compared to global PC expansions. We propose a methodology that first finds a probabilistic description of the discontinuity given a number of data points. Assuming the discontinuity curve is a polynomial, the algorithm is based on Bayesian inference of its coefficients. Markov chain Monte Carlo sampling is used to obtain joint distributions for the polynomial coefficients, effectively parameterizing the distribution over all possible discontinuity curves. Next, we apply the Rosenblatt transformation to the irregular parameter domains on each side of the discontinuity. This transformation maps a space of uncertain parameters with specific probability distributions to a space of i.i.d standard random variables where orthogonal projections can be used to obtain PC coefficients. In particular, we use uniform random variables that are compatible with PC expansions based on Legendre polynomials. The Rosenblatt transformation and the corresponding PC expansions for the model output on either side of the discontinuity are applied successively for several realizations of the discontinuity curve. The climate model output and its associated uncertainty at specific design points is then computed by taking a quadrature-based integration average over PC expansions corresponding to possible realizations of the discontinuity curve.

54 ENVIRONMENTAL SCIENCES↗

Comparing probabilistic and descriptive analyses of time–dose–toxicity relationship for determining no-observed-adverse-effect level in drug development

The no-observed-adverse-effect level (NOAEL) of a drug defined from animal studies is important for inferring a maximal safe dose in human. However, several issues are associated with its concept, determination and application. It is confined to the actual doses used in the study; becomes lower with increasing sample size or dose levels; and reflects the risk level seen in the experiment rather than what may be relevant for human. We explored a pharmacometric approach in an attempt to address these issues. We first used simulation to examine the behaviour of the NOAEL values as determined by current common practice; and then fitted the probability of toxicity as a function of treatment duration and dose to data collected from all applicable toxicology studies of a test compound. Our investigation was in the context of an irreversible toxicity that is detected at the end of the study. Simulations illustrated NOAEL's dependency on experimental factors such as dose and sample size, as well as the underlying uncertainty. Modelling the probability as a continuous function of treatment duration and dose simultaneously to data from multiple studies allowed the estimation of the dose, along with its confidence interval, for a maximal risk level that might be deemed as acceptable for human. The model-based data integration also reconciled between-study inconsistency and explicitly provided maximised estimation confidence. Such alternative NOAEL determination method should be explored for its more efficient data use, more quantifiable insight to toxic doses, and the potential for more relevant animal-to-human translation. - Highlights: • Simulations revealed issues with NOAEL concept, determination and application. • Probabilistic modelling was used to address these issues. • The model integrated time-dose-toxicity data from multiple studies. • The approach uses data efficiently and may allow more meaningful human translation.

60 APPLIED LIFE SCIENCES↗

NESSUS/expert and NESSUS/FPI in the Probabilistic Structural Analysis Methods (PSAM) program

The Numerical Evaluation of Stochastic Structures under Stress (NESSUS) is the primary computer code being developed in the NASA Probabilistic Structural Analysis Methods (PSAM) project. It consists of four modules NESSUS/EXPERT, NESSUS/FPI, NESSUS/PRE and NESSUS/FEM. This presentation concentrates on EXPERT and FPI. To provide an effective interface between NESSUS and the user, an expert system module called NESSUS/EXPERT is being developed. That system uses the CLIPS artificial intelligence code developed to NASA-JSC. The code is compatible with FORTRAN, the standard language for codes in PSAM. The user interacts with the CLIPS inference engine, which is linked to the knowledge database. The perturbation database generated by NESSUS/FEM and managed in EXPERT is used to develop the so-called response or performance model in the random variables. Two independent probabilistic methods are available in PSAM for the computation of the probabilistic structural response. These are the Fast Probability Integration (FPI) method and Monte Carlo simulation. FPI is classified as an advanced reliability method and has been developed over the past ten years by researchers addressing the reliability of civil engineering structures. Monte Carlo is a well-established technique for computing probabilities by conducting a number of deterministic analyses with specified input distributional information.

Burnside, O. H.↗

ENSEMBLE KALMAN FILTERS FOR RELIABILITY ESTIMATION IN PERFUSION INFERENCE

We consider the solution of inverse problems in dynamic contrast–enhanced imaging by means of ensemble Kalman filters. Our quantity of interest is blood perfusion, i.e., blood flow rates in tissue. While existing approaches to compute blood perfusion parameters for given time series of radiological measurements mainly rely on deterministic, deconvolution–based methods, we aim at recovering probabilistic solution information for given noisy measurements. To this end, we model radiological image capturing as a sequential data assimilation process and solve it by an ensemble Kalman filter. Thereby, we recover deterministic results as an ensemble–based mean and are able to compute reliability information such as probabilities for the perfusion to be in a given range. Our target application is the inference of blood perfusion parameters in the human brain. A numerical study shows promising results for artificial measurements generated by a digital perfusion phantom.

Engineering↗

Statistical Treatment of Inverse Problems Constrained by Differential Equations-based Models with Stochastic Terms

This paper introduces a statistical treatment of inverse problems constrained by models with stochastic terms. The solution of the forward problem is given by a distribution represented numerically by an ensemble of simulations. The goal is to formulate the inverse problem, in particular the objective function, to find the closest forward distribution (i.e., the output of the stochastic forward problem) that best explains the distribution of the observations in a certain metric. We use proper scoring rules, a concept employed in statistical forecast verification, namely energy, variogram, and hybrid (i.e., combination of the two) scores. We study the performance of the proposed formulation in the context of two applications: a coefficient field inversion for subsurface flow governed by an elliptic partial differential equation with a stochastic source and a parameter inversion for power grid governed by differential-algebraic equations. In both cases we show that the variogram and the hybrid scores produce better parameter inversion results than does the energy score, whereas the energy score leads to better probabilistic predictions.

97 MATHEMATICS AND COMPUTING↗

A Bayesian Approach to Real-Time Dynamic Parameter Estimation Using Phasor Measurement Unit Measurement

Here, we develop a polynomial-chaos-expansion (PCE)-based approach for decentralized dynamic parameter estimation through Bayesian inference. Using this approach, the non-Gaussian distribution of the inverted parameters is obtained. More specifically, we first represent the decentralized generator model with the PCE-based surrogate. This surrogate allows us to efficiently evaluate the time-consuming dynamic solver at parameter values through Metropolis-Hastings (M-H)-based Markov chain Monte Carlo (MCMC). Then, we propose a two-stage hybrid Markov chain Monte Carlo (MCMC) to recover a posteriori distribution of the decentralized generator model parameters. In the first stage, we use the gradient-enhanced Langevin MCMC algorithm to characterize an intermediate posterior parameter distribution. This algorithm is computationally scalable to the high-dimensional parameter space. Based on the intermediate posterior distribution, during the second stage, we use the adaptive MCMC algorithm to fine-tune the strong correlations between the parameters. Finally, the fully recovered a posterior distribution is obtained in the end. The simulation results show that the proposed PCE-based hybrid MCMC algorithm can accurately and efficiently estimate the high-dimensional generator dynamic model parameters with full probabilistic distribution provided.

42 ENGINEERING↗

Learning from crowds with variational Gaussian processes

Solving a supervised learning problem requires to label a training set. This task is traditionally performed by an expert, who provides a label for each sample. The proliferation of social web services (e.g., Amazon Mechanical Turk) has introduced an alternative crowdsourcing approach. Anybody with a computer can register in one of these services and label, either partially or completely, a dataset. The effort of labeling is then shared between a great number of annotators. However, this approach introduces scientifically challenging problems such as combining the unknown expertise of the annotators, handling disagreements on the annotated samples, or detecting the existence of spammer and adversarial annotators. All these problems require probabilistic sound solutions which go beyond the naive use of majority voting plus classical classification methods. In this work we introduce a new crowdsourcing model and inference procedure which trains a Gaussian Process classifier using the noisy labels provided by the annotators. Variational Bayes inference is used to estimate all unknowns. The proposed model can predict the class of new samples and assess the expertise of the involved annotators. Moreover, the Bayesian treatment allows for a solid uncertainty quantification. Since when predicting the class of a new sample we might have access to some annotations for it, we also show how our method can naturally incorporate this additional information. Furthermore, a comprehensive experimental section evaluates the proposed method with synthetic and real experiments, showing that it consistently outperforms other state-of-the-art crowdsourcing approaches.

97 MATHEMATICS AND COMPUTING↗

Aviation Safety Risk Modeling: Lessons Learned From Multiple Knowledge Elicitation Sessions

Aviation safety risk modeling has elements of both art and science. In a complex domain, such as the National Airspace System (NAS), it is essential that knowledge elicitation (KE) sessions with domain experts be performed to facilitate the making of plausible inferences about the possible impacts of future technologies and procedures. This study discusses lessons learned throughout the multiple KE sessions held with domain experts to construct probabilistic safety risk models for a Loss of Control Accident Framework (LOCAF), FLightdeck Automation Problems (FLAP), and Runway Incursion (RI) mishap scenarios. The intent of these safety risk models is to support a portfolio analysis of NASA's Aviation Safety Program (AvSP). These models use the flexible, probabilistic approach of Bayesian Belief Networks (BBNs) and influence diagrams to model the complex interactions of aviation system risk factors. Each KE session had a different set of experts with diverse expertise, such as pilot, air traffic controller, certification, and/or human factors knowledge that was elicited to construct a composite, systems-level risk model. There were numerous "lessons learned" from these KE sessions that deal with behavioral aggregation, conditional probability modeling, object-oriented construction, interpretation of the safety risk results, and model verification/validation that are presented in this paper.

Luxhoj, J. T.↗

Environmental Selection, Dispersal, and Organism Interactions Shape Community Assembly in High-Throughput Enrichment Culturing

A central goal of microbial ecology is to identify and quantify the forces that lead to observed population distributions and dynamics. However, these forces, which include environmental selection, dispersal, and organism interactions, are often difficult to assess in natural environments. Here in this paper, we present a method that links microbial community structures with selective and stochastic forces through highly replicated subsampling and enrichment of a single environmental inoculum. Specifically, groundwater from a well-studied natural aquifer was serially diluted and inoculated into nearly 1,000 aerobic and anaerobic nitrate-reducing cultures, and the final community structures were evaluated with 16S rRNA gene amplicon sequencing. We analyzed the frequency and abundance of individual operational taxonomic units (OTUs) to understand how probabilistic immigration, relative fitness differences, environmental factors, and organismal interactions contributed to divergent distributions of community structures. We further used a most probable number (MPN) method to estimate the natural condition-dependent cultivable abundance of each of the nearly 400 OTU cultivated in our study and infer the relative fitness of each. Additionally, we infer condition-specific organism interactions and discuss how this high-replicate culturing approach is essential in dissecting the interplay between overlapping ecological forces and taxon-specific attributes that underpin microbial community assembly. IMPORTANCE Through highly replicated culturing, in which inocula are subsampled from a single environmental sample, we empirically determine how selective forces, interspecific interactions, relative fitness, and probabilistic dispersal shape bacterial communities. These methods offer a novel approach to untangle not only interspecific interactions but also taxon-specific fitness differences that manifest across different cultivation conditions and lead to the selection and enrichment of specific organisms. Additionally, we provide a method for estimating the number of cultivable units of each OTU in the original sample through the MPN approach.

16S RNA↗

Physics-Informed Sparse Gaussian Process for Probabilistic Stability Analysis of Large-Scale Power System with Dynamic PVs and Loads

This work proposes a physics-informed sparse Gaussian process (SGP) for probabilistic stability assessment of large-scale power systems in the presence of uncertain dynamic PVs and loads. The differential and algebraic equations considering uncertainties from dynamic PVs and loads are reformulated to a nonlinear mapping relationship that allows the application of SGP. Thanks to the nonparametric characteristic of Gaussian process, the proposed framework does not require distributions of uncertain inputs and this distinguishes it from existing approaches. As the original Gaussian process is not scalable to large-scale systems with high dimensional uncertain inputs, this paper develops the SGP with a stochastic variational inference technique. It leads to approximately two orders of complex reduction. A data pre-processing step is also introduced to tackle the coexistence of stable and unstable cases by sample clustering and constructing separate SGPs. The probabilistic transient stability index is analyzed to assess system stability under different uncertain dynamics loads and PVs. Comparisons are performed with the sampling-based, the polynomial chaos expansion-based, and traditional Gaussian process-based methods on the modified IEEE 118-bus and Texas 2000-bus systems under various scenarios, including different levels of uncertainties and the existence of nonlinear correlations among dynamic PVs. The impacts of data quality and quantity issues are also investigated. It is shown that the proposed SGP achieves significantly improved computational efficiency while maintaining high accuracy with a limited number of data.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Leveraging Structured Biological Knowledge for Counterfactual Inference: A Case Study of Viral Pathogenesis

Counterfactual inference is a useful tool for comparing outcomes of interventions on complex systems. It requires us to represent the system in form of a structural causal model, complete with a causal diagram, probabilistic assumptions on exogenous variables, and functional assignments. Specifying such models can be extremely difficult in practice. The process requires substantial domain expertise, and does not scale easily to large systems, multiple systems, or novel system modifications. At the same time, many application domains, such as molecular biology, are rich in structured causal knowledge that is qualitative in nature. This manuscript proposes a general approach for querying a causal knowledge graph with a causal question and converting the qualitative result into a quantitative structural causal model that can learn from data to answer the question. Here, we demonstrate the feasibility, accuracy and versatility of this approach using two case studies in systems biology. The first demonstrates the appropriateness of the underlying assumptions and the accuracy of the results. The second demonstrates the versatility of the approach by querying a knowledge base for the molecular determinants of a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-induced cytokine storm and performing counterfactual inference to predict the causal effect of medical countermeasures for severely ill COVID-19 patients.

60 APPLIED LIFE SCIENCES↗

A regression-based approach to estimating retrofit savings using the Building Performance Database

Retrofitting building systems is known to provide cost-effective energy savings. This article addresses how the Building Performance Database is used to help identify potential savings. Currently, prioritizing retrofits and computing their expected energy savings and cost/benefits can be a complicated, costly, and an uncertain effort. Prioritizing retrofits for a portfolio of buildings can be even more difficult if the owner must determine different investment strategies for each of the buildings. Meanwhile, we are seeing greater availability of data on building energy use, characteristics, and equipment. These data provide opportunities for the development of algorithms that link building characteristics and retrofits empirically. In this paper we explore the potential of using such data for predicting the expected energy savings from equipment retrofits for a large number of buildings. We show that building data with statistical algorithms can provide savings estimates when detailed energy audits and physics-based simulations are not cost- or time-feasible. We develop a multivariate linear regression model with numerical predictors (e.g., operating hours, occupant density) and categorical indicator variables (e.g., climate zone, heating system type) to predict energy use intensity. The model quantifies the contribution of building characteristics and systems to energy use, and we use it to infer the expected savings when modifying particular equipment. We verify the model using residual analysis and cross-validation. We demonstrate the retrofit analysis by providing a probabilistic estimate of energy savings for several hypothetical building retrofits. We discuss the ways understanding the risk associated with retrofit investments can inform decision making. The contributions of this work are the development of a statistical model for estimating energy savings, its application to a large empirical building dataset, and a discussion of its use in informing building retrofit decisions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Using ultrasonic attenuation in cortical bone to infer distributions on pore size

Here, in this work we infer the underlying distribution on pore radius in human cortical bone samples using ultrasonic attenuation data. We first discuss how to formulate polydisperse attenuation models using a probabilistic approach and the Waterman Truell model for scattering attenuation. We then compare the Independent Scattering Approximation and the higher-order Waterman Truell models’ forward predictions for total attenuation in polydisperse samples. Following this, we formulate an inverse problem under the Prohorov Metric Framework coupled with variational regularization to stabilize this inverse problem. We then use experimental attenuation data taken from human cadaver samples and solve inverse problems resulting in nonparametric estimates of the probability density function on pore radius. We compare these estimates to the “true” microstructure of the bone samples determined via microCT imaging. We find that our methodology allows us to reliably estimate the underlying microstructure of the bone from attenuation data.

97 MATHEMATICS AND COMPUTING↗