Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “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 271 records · Page 15

Approximate inference for constructing astronomical catalogs from images

Here, we present a new, fully generative model for constructing astronomical catalogs from optical telescope image sets. Each pixel intensity is treated as a random variable with parameters that depend on the latent properties of stars and galaxies. These latent properties are themselves modeled as random. We compare two procedures for posterior inference. One procedure is based on Markov chain Monte Carlo (MCMC) while the other is based on variational inference (VI). The MCMC procedure excels at quantifying uncertainty, while the VI procedure is 1000 times faster. On a supercomputer, the VI procedure efficiently uses 665,000 CPU cores to construct an astronomical catalog from 50 terabytes of images in 14.6 minutes, demonstrating the scaling characteristics necessary to construct catalogs for upcoming astronomical surveys.

59 BASIC BIOLOGICAL SCIENCES↗

Exploring the physics of cesium lead halide perovskite quantum dots via Bayesian inference of the photoluminescence spectra in automated experiment

The unique optoelectronic properties of metal halide perovskite quantum dots (QDs) make them promising candidates for applications in light-emitting diodes (LEDs), scintillators, and other photonic devices. The automated micropipetting synthesis platform equipped with an optical reader enables the opportunity for high throughput synthesis and photoluminescent (PL) characterization of metal halide perovskite QDs for the first time. Here, we explore the compositional dependence of the PL behavior and stability of the combinatorial library of cesium lead halide (CsPbX 3 ) perovskites QDs via the automated platform. To study the stability of synthesized QDs in the binary and ternary configurations, we study the time-dependent PL properties using previously developed machine learning analysis. Additionally, to systematically explore the PL behavior in the ternary CsPbX 3 QDs system, we introduce the Bayesian inference framework that allows the probabilistic fit of multiple models to the PL data and establishes both optimal model and model parameter robustly. Furthermore, these behaviors can be used as a control parameter for the navigation of the multidimensional compositional spaces in automated synthesis. Furthermore, this analysis shows the nonuniformity of the PL peak behavior in the ternary CsPbX 3 QDs system. Further, the analysis confirms narrow size distribution and good quality of CsPbBr 3 QDs alloyed with low concentrations of iodide and chloride. We note that Bayesian Inference fit parameters can be further used as a control signal for navigation of the chemical spaces in automated synthesis.

36 MATERIALS SCIENCE↗

Model Structural Inference Using Local Dynamic Operators

This paper focuses on the problem of quantifying the effects of model-structure uncertainty in the context of time-evolving dynamical systems. This is motivated by multi-model uncertainty in computer physics simulations: developers often make different modeling choices in numerical approximations and process simplifications, leading to different numerical codes that ostensibly represent the same underlying dynamics. We consider model-structure inference as a two-step methodology: the first step is to perform system identification on numerical codes for which it is possible to observe the full state; the second step is structural uncertainty quantification, in which the goal is to search candidate models "close" to the numerical code surrogates for those that best match a quantity of interest (QOI) from some empirical data sets. Here, we (1) define a discrete, local representation of the structure of a partial differential equation, which we refer to as the "local dynamical operator" (LDO); (2) identify model structure nonintrusively from numerical code output; (3) nonintrusively construct a reduced-order model (ROM) of the numerical model through POD-DEIM-Galerkin projection; (4) perturb the ROM dynamics to approximate the behavior of alternate model structures; and (5) apply Bayesian inference and energy conservation laws to calibrate a LDO to a given QOI. We demonstrate these techniques using the two-dimensional rotating shallow water equations as an example system.

97 MATHEMATICS AND COMPUTING↗

Parsimonious Inference Information-Theoretic Foundations for a Complete Theory of Machine Learning (CIS-LDRD Project 218313 Final Technical Report)

This work examines how we may cast machine learning within a complete Bayesian framework to quantify and suppress explanatory complexity from first principles. Our investigation into both the philosophy and mathematics of rational belief leads us to emphasize the critical role of Bayesian inference in learning well-justified predictions within a rigorous and complete extended logic. The Bayesian framework allows us to coherently account for evidence in the learned plausibility of potential explanations. As an extended logic, the Bayesian paradigm regards probability as a notion of degrees of truth. In order to satisfy critical properties of probability as a coherent measure, as well as maintain consistency with binary propositional logic, we arrive at Bayes' Theorem as the only justifiable mechanism to update our beliefs to account for empiracle evidence. Yet, in the machine learning paradigm, where explanations are unconstrained algorithmic abstractions, we arrive at a critical challenge: Bayesian inference requires prior belief. Conventional approaches fail to yield a consistent framework in which we could compare prior plausibility among the infinities of potential choices in learning architectures. The difficulty of articulating well-justified prior belief over abstract models is the provinence of memorization in traditional machine learning training practices. This becomes exceptionally problematic in the context of limited datasets, when we wish to learn justifiable predictions from only a small amount of data.

97 MATHEMATICS AND COMPUTING↗

Industrial PLC Network Modeling and Parameter Identification Using Sensitivity Analysis and Mean Field Variational Inference

A multiple input multiple output (MIMO) power line communication (PLC) model for industrial facilities was developed that uses the physics of a bottom-up model but can be calibrated like top-down models. The PLC model considers 4-conductor cables (three-phase conductors and a ground conductor) and has several load types, including motor loads. The model is calibrated to data using mean field variational inference with a sensitivity analysis to reduce the parameter space. The results show that the inference method can accurately identify many of the model parameters, and the model is accurate even when the network is modified.

47 OTHER INSTRUMENTATION↗

Parameter inference with deep jointly informed neural networks

Abstract A common challenge in modeling inertial confinement fusion (ICF) experiments with computer simulations is that many of the simulation inputs are unknown and cannot be directly measured. Often, parameters that are measured in the experiment are used to infer the unknown inputs by solving the inverse problem: finding the set of simulation inputs that result in outputs consistent with the experimental observations. In ICF, this process is often referred to as a “post‐shot analysis.” Post‐shot analyses are challenging as the inverse problem is often highly degenerate, the input parameter space is vast, and simulations are computationally expensive. In this work, deep neural network models equipped with model uncertainty estimates are used to train inverse models, which map directly from output to input space, to find the distribution of post‐shot simulations that are consistent with experimental observations. The inverse model approach is compared to Markov chain Monte Carlo (MCMC) sampling of the forward model, which maps from input to output space, for parameter inference tasks of varying complexity. The inverse models perform best when searching vast parameter spaces for post‐shot simulations that are consistent with a large number of observables, where MCMC sampling can be prohibitively expensive. We demonstrate how augmenting inverse models with autoencoders enable the inclusion of several dozen observables in the inverse mapping, reducing the degeneracy of the model and improving the accuracy of the post‐shot analysis.

Humbird, Kelli D.↗

Data mining and statistical inference in selective laser melting

Selective laser melting (SLM) is an additive manufacturing process that builds a complex three-dimensional part, layer-by-layer, using a laser beam to fuse fine metal powder together. The design freedom afforded by SLM comes associated with complexity. As the physical phenomena occur over a broad range of length and time scales, the computational cost of modeling the process is high. At the same time, the large number of parameters that control the quality of a part make experiments expensive. In this paper, we describe ways in which we can use data mining and statistical inference techniques to intelligently combine simulations and experiments to build parts with desired properties. We start with a brief summary of prior work in finding process parameters for high-density parts. We then expand on this work to show how we can improve the approach by using feature selection techniques to identify important variables, data-driven surrogate models to reduce computational costs, improved sampling techniques to cover the design space adequately, and uncertainty analysis for statistical inference. Here, our results indicate that techniques from data mining and statistics can complement those from physical modeling to provide greater insight into complex processes such as selective laser melting.

36 MATERIALS SCIENCE↗

Linear two-pool models are insufficient to infer soil organic matter decomposition temperature sensitivity from incubations

Abstract Terrestrial carbon (C)-climate feedbacks depend strongly on how soil organic matter (SOM) decomposition responds to temperature. This dependency is often represented in land models by the parameter Q 10 , which quantifies the relative increase of microbial soil respiration per 10 °C temperature increase. Many studies have conducted paired laboratory soil incubations and inferred “active” and “slow” pool Q 10 values by fitting linear two-pool models to measured respiration time series. Using a recently published incubation study (Qin et al. in Sci Adv 5(7):eaau1218, 2019) as an example, here we first show that the very high parametric equifinality of the linear two-pool models may render such incubation-based Q 10 estimates unreliable. In particular, we show that, accompanied by the uncertain initial active pool size, the slow pool Q 10 can span a very wide range, including values as high as 100, although all parameter combinations are producing almost equally good model fit with respect to the observations. This result is robust whether or not interactions between the active and slow pools are considered (typically these interactions are not considered when interpreting incubation data, but are part of the predictive soil carbon models). This very large parametric equifinality in the context of interpreting incubation data is consistent with the poor temporal extrapolation capability of linear multi-pool models identified in recent studies. Next, using a microbe-explicit SOM model (RESOM), we show that the inferred two pools and their associated parameters (e.g., Q 10 ) could be artificial constructs and are therefore unreliable concepts for integration into predictive models. We finally discuss uncertainties in applying linear two-pool (or more generally multiple-pool) models to estimate SOM decomposition parameters such as temperature sensitivities from laboratory incubations. We also propose new observations and model structures that could enable better process understanding and more robust predictive capabilities of soil carbon dynamics.

54 ENVIRONMENTAL SCIENCES↗

Accelerating Machine Learning Inference with GPUs in ProtoDUNE Data Processing

Abstract We study the performance of a cloud-based GPU-accelerated inference server to speed up event reconstruction in neutrino data batch jobs. Using detector data from the ProtoDUNE experiment and employing the standard DUNE grid job submission tools, we attempt to reprocess the data by running several thousand concurrent grid jobs, a rate we expect to be typical of current and future neutrino physics experiments. We process most of the dataset with the GPU version of our processing algorithm and the remainder with the CPU version for timing comparisons. We find that a 100-GPU cloud-based server is able to easily meet the processing demand, and that using the GPU version of the event processing algorithm is two times faster than processing these data with the CPU version when comparing to the newest CPUs in our sample. The amount of data transferred to the inference server during the GPU runs can overwhelm even the highest-bandwidth network switches, however, unless care is taken to observe network facility limits or otherwise distribute the jobs to multiple sites. We discuss the lessons learned from this processing campaign and several avenues for future improvements.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Detecting dynamic causal inference in nonlinear two-phase fracture flow

Identifying dynamic causal inference involved in flow and transport processes in complex fractured-porous media is generally a challenging task, because nonlinear and chaotic variables may be positively coupled or correlated for some periods of time, but can then become spontaneously decoupled or non-correlated. In his 2002 paper (Faybishenko, 2002), the author performed a nonlinear dynamical and chaotic analysis of time-series data obtained from the fracture flow experiment conducted by Persoff and Pruess (1995), and, based on the visual examination of time series data, hypothesized that the observed pressure oscillations at both inlet and outlet edges of the fracture result from a superposition of both forward and return waves of pressure propagation through the fracture. In the current paper, the author explores an application of a combination of methods for detecting nonlinear chaotic dynamics behavior along with the multivariate Granger Causality (G-causality) time series test. Based on the G-causality test, the author infers that his hypothesis is correct, and presents a causation loop diagram of the spatial-temporal distribution of gas, liquid, and capillary pressures measured at the inlet and outlet of the fracture. The causal modeling approach can be used for the analysis of other hydrological processes, for example, infiltration and pumping tests in heterogeneous subsurface media, and climatic processes, for example, to find correlations between various meteorological parameters, such as temperature, solar radiation, barometric pressure, etc.

54 ENVIRONMENTAL SCIENCES↗

A large-scale evaluation of automated metadata inference approaches on sensors from air handling units

Building automation systems provide abundant sensor data to enable the potential of using data analytics to, among other things, improve the energy efficiency of the building. However, deployment of these applications for buildings, such as, fault detection and diagnosis (FDD) on multiple buildings remains a challenge due to the non-trivial efforts of organizing, managing and extracting metadata associated with sensors (e.g., information about their location, function, etc.), which is required by applications. One of the reasons leading to the problem is that varying conventions, acronyms, and standards are used to define this metadata. To better understand the nature of the problem, as well as the performance and scalability of existing solutions, in this report we implement and test 6 different time-series based metadata inference approaches on sensors from 614 air handling units (AHU) instrumented in 35 building sites accounting for more than 400 buildings distributed across United States of America. We infer 12 types of sensors and actuators in AHUs required by a rule-based FDD application: AHU performance and assessment rules (APAR). Our results show that: (1) the average performance of these approaches in terms of accuracy is similar across building sites, though there is significant variance; (2) the expected accuracy of classifying the type of points required by APAR for a new unseen building is, on average, 75%; (3) the performance of the model does not decrease as long as training data and testing data are extracted from adjacent months.

42 ENGINEERING↗

Inferring occupant counts from Wi-Fi data in buildings through machine learning

An important approach to curtail building energy consumption is to optimize building control based on occupancy information. Various studies proposed to estimate occupant counts through different approaches and sensors. However, high cost and privacy concerns remain as major barriers, restricting the practice of occupant count detection. In this study, we propose a novel method utilizing data from widely deployed Wi-Fi infrastructure to infer occupant counts through machine learning. Compared with the current indirect measurement methods, our method improves the performance of estimating people count: (1) we avoid privacy concerns by anonymizing and reshuffling the MAC addresses on a daily basis; (2) we adopted a heuristic feature engineer approach to cluster connected devices into different types based on their daily connection duration. We tested the method in an office building located in California. In an area with an average occupancy of 22–27 people and a peak occupancy of 48 - 74 people, the root square mean error on the test set is less than four people. The error is within two people counts for more than 70% of estimations, and less than six counts for more than 90% of estimations, indicating a relatively high accuracy. The major contribution of this study is proposing a novel and accurate approach to detect occupant counts in a non-intrusive way, i.e., utilizing existing Wi-Fi infrastructure in buildings without requiring the installation of extra hardware or sensors. The method we proposed is generic and could be applied to other commercial buildings to infer occupant counts for energy efficient building control.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Onset of magma ocean solidification on Mars inferred from Mn-Cr chronometry

The mantle of Mars probably differentiated through the crystallization of a magma ocean during the first tens of million years (Ma) of Solar System evolution. However, the exact timescale of large-scale silicate differentiation of the martian mantle is debated, and in particular, it remains unclear when differentiation commenced. In this paper, we applied the short-lived 53 Mn- 53 Cr system to martian meteorites in order to date the onset of large-scale mantle differentiation on Mars. The new Cr isotope data demonstrate that martian meteorites exhibit no resolvable radiogenic 53 Cr variations, and instead have a uniform +20.2±1.2 (95% conf.) parts-per-million excess in 53 Cr/ 52 Cr relative to the terrestrial mantle. The investigated groups of martian meteorites are lithologically varied and derive from diverse mantle sources that probably had variable Mn/Cr. Hence, the lack of 53 Cr variability among martian meteorites demonstrates that silicate differentiation on Mars occurred after the extinction of 53 Mn. Provided that the sources of the martian meteorites have Mn/Cr variations that are typical of the terrestrial planets, this result implies that the onset of large-scale silicate differentiation must have occurred later than 20±5 Ma after Solar System formation. The onset of silicate differentiation on Mars inferred here is significantly later than time estimates for segregation of the martian core which conservatively occurred within <10 Ma after Solar System formation. Thus, the new Mn-Cr data imply that there was a small, but resolvable, time gap of at least 5 Ma between core formation and magma ocean solidification on Mars. If the age of core segregation is taken at face value, our results imply that the martian magma ocean remained mostly molten over several Ma. This inferred longevity of the magma ocean is inconsistent with thermal models predicting rapid (<1 Ma) solidification of the martian magma ocean. Although there is currently no unique solution to this conundrum, our results can potentially be explained by a protracted history of impact bombardment that delayed differentiation in a shallow magma ocean on Mars, or perhaps more readily, by the presence of an early and dense atmosphere that acted as an insulator and prevented the magma ocean from cooling quickly.

58 GEOSCIENCES↗

Generalized fiducial inference on differentiable manifolds

We introduce a novel approach to inference on parameters that take values in a Riemannian manifold embedded in a Euclidean space. Parameter spaces of this form are ubiquitous across many fields, including chemistry, physics, computer graphics, and geology. Here, this new approach uses generalized fiducial inference (GFI) to obtain a posterior-like distribution on the manifold, without needing to know local parameterizations that map to the constrained space from an unconstrained Euclidean space. Using mathematical tools from Riemannian geometry, we construct a constrained generalized fiducial distribution (CGFD). A Bernstein-von Mises-type result for the CGFD, which provides intuition for how the desirable asymptotic qualities of the unconstrained generalized fiducial distribution are inherited by the CGFD, is provided. To illustrate the practical use of the CGFD, we provide a proof-of-concept example in the context of a linear logspline density estimation problem, and demonstrate that CGFD-based confidence sets exhibit desirable coverage properties via simulation. As an application, we fit a CGFD to COVID-19 case count data from North Carolina, USA.

97 MATHEMATICS AND COMPUTING↗

A Bayesian statistical approach to inferring particle dynamics from in-situ pump POC and chloropigment data from the Mediterranean Sea

Concentrations of chloropigments and particulate organic carbon (POC) in large-volume in-situ pump samples from the Mediterranean Sea were used to estimate rate constants of processes that control the fate of particles, and specifically chloropigments, in the water column. In this paper we introduce a Bayesian statistical inversion method that combines the data with a new box model and has the capacity to infer rate constants for POC respiration/dissolution, chlorophyll and pheopigment degradation, and particle aggregation and disaggregation. We use first-order kinetics to model disaggregation, and use both first-order and second-order kinetics to model aggregation. Using these methods, the estimated small-particle (1–70 μm) POC respiration rate constant was 2.44 -1.00 +1.69 yr -1 (0.41 yr). The estimated disaggregation and second-order aggregation rate constants were 85.6 -36.4 +63.4 yr -1 (1.17×10 -2 yr) and 2.78 -1.17 +2.01 μM -1 yr -1 , respectively. Using the optimal rate constants and the corresponding particle concentrations, disaggregation is ~4.2 times faster than the small-size POC dissolution rate, which indicates that disaggregation is a dominant processes at the time of sampling. More importantly, by comparing our results with those of previous studies, we conclude that sampling methods have less influence than tracers themselves on inferring particle dynamic rate constants. We previously introduced a somewhat similar approach to modeling SV sediment trap data, but large volume pumps are a much more common sample collection method in oceanographic surveys than SV sediment traps, and thus our new model should have a wider applicability.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

An iterative learning and inference approach to managing dynamic cyber-vulnerabilities of complex systems

As modern systems become increasingly reliant on cyber-technologies and continue to be integrated with physical systems, managing risks from deliberate and non-deliberate sources is a significant research challenge. Unlike strictly physical systems, cyber-enabled physical systems are influenced by dynamic and evolving technologies, environments, and attack mechanisms. As a result, vulnerabilities are rapidly changing and difficult to detect and manage. While there is recent interest in the dynamic properties of performance through resilience analysis, limited research addresses the dynamic nature of cyber-system vulnerability. This paper presents a dynamic iterative learning approach to evaluate the state of health for systems as related to cyber-vulnerabilities. These time-varying system health characteristics may not be visible, but can be inferred using observable characteristics. The approach recognizes that multiple types of vulnerabilities need to be included in a holistic system health assessment. The methods are applied to the Common Vulnerability Scoring System database containing over 100,000 documented cybersecurity vulnerabilities. This work is the first to acknowledge the dynamic properties of cyber-vulnerability, while also inferring system health using observable data and hidden states of system health. This work will be of interest to managers of large-scale cyber-enabled physical systems who are seeking to prioritize system health investments.

Chatterjee, Samrat↗

A Bayesian model for multivariate discrete data using spatial and expert information with application to inferring building attributes

When modeling sparsely observed multivariate data, strong prior information elicited from experts can be used to bolster predictive accuracy and counteract sampling bias. Similarly, modeling autocorrelation in space can help make use of co-occurrence patterns present in many types of spatial data. To make use of both expert prior information and spatial structure, we propose a novel graphical model for a spatial Bayesian network developed specifically to address challenges in inferring the attributes of buildings from geographically sparse observational data. This model is implemented as the sum of a spatial multivariate Gaussian random field and a tabular conditional probability function in real-valued space prior to projection onto the probability simplex. This modeling form is especially suitable for the usage of prior information in the form of sets of atomic rules obtained from experts. To perform inference with missing data, we implement a Markov chain Monte Carlo scheme composed of alternating steps of Gibbs sampling of missing entries and Hamiltonian Monte Carlo for model parameters. A case study in building attribution is presented to highlight the advantages and limitations of this approach.

97 MATHEMATICS AND COMPUTING↗

A survey of techniques for optimizing transformer inference

Recent years have seen a phenomenal rise in the performance and applications of transformer neural networks. The family of transformer networks, including Bidirectional Encoder Representations from Transformer (BERT), Generative Pretrained Transformer (GPT) and Vision Transformer (ViT), have shown their effectiveness across Natural Language Processing (NLP) and Computer Vision (CV) domains. Transformer-based networks such as ChatGPT have impacted the lives of common men. However, the quest for high predictive performance has led to an exponential increase in transformers' memory and compute footprint. Researchers have proposed techniques to optimize transformer inference at all levels of abstraction. Further, this paper presents a comprehensive survey of techniques for optimizing the inference phase of transformer networks. We survey techniques such as knowledge distillation, pruning, quantization, neural architecture search and lightweight network design at the algorithmic level. We further review hardware-level optimization techniques and the design of novel hardware accelerators for transformers. We summarize the quantitative results on the number of parameters/FLOPs and the accuracy of several models/techniques to showcase the tradeoff exercised by them. We also outline future directions in this rapidly evolving field of research. We believe that this survey will educate both novice and seasoned researchers and also spark a plethora of research efforts in this field.

97 MATHEMATICS AND COMPUTING↗