Results for “hierarchical models” — Page 11

DOE OSTI2021

Bridging Multiscale Processes in Earth System Models with Physics-Guided Hierarchical Machine Learning

Focal Area(s): The focal area of this whitepaper is, “Predictive modeling through the use of AI techniques and AI-derived model components; the use of AI and other tools to design a prediction system comprising of a hierarchy of models (e.g., AI driven model/component/parameterization selection).” The ideas and frameworks described herein are deemed site agnostic. As a use case, this group will initially focus on the coupling between land processes, surface/subsurface hydrological processes, coastal processes, and human activities in the U.S. Gulf states.

54 ENVIRONMENTAL SCIENCES↗

NASA NTRS1995

Hierarchical Feedback Decomposition of Steady-State Equation Models

Explore the source record for details and available documents.

Hierarchical Feedback↗

NASA NTRS1991

Using GOMS models and hypertext to create representations of medical procedures for online display

This study investigated two methods to improve organization and presentation of computer-based medical procedures. A literature review suggested that the GOMS (goals, operators, methods, and selecton rules) model can assist in rigorous task analysis, which can then help generate initial design ideas for the human-computer interface. GOMS model are hierarchical in nature, so this study also investigated the effect of hierarchical, hypertext interfaces. We used a 2 x 2 between subjects design, including the following independent variables: procedure organization - GOMS model based vs. medical-textbook based; navigation type - hierarchical vs. linear (booklike). After naive subjects studies the online procedures, measures were taken of their memory for the content and the organization of the procedures. This design was repeated for two medical procedures. For one procedure, subjects who studied GOMS-based and hierarchical procedures remembered more about the procedures than other subjects. The results for the other procedure were less clear. However, data for both procedures showed a 'GOMSification effect'. That is, when asked to do a free recall of a procedure, subjects who had studies a textbook procedure often recalled key information in a location inconsistent with the procedure they actually studied, but consistent with the GOMS-based procedure.

Gugerty, Leo↗

NASA NTRS1987

A qualitative model of human interaction with complex dynamic systems

A qualitative model describing human interaction with complex dynamic systems is developed. The model is hierarchical in nature and consists of three parts: a behavior generator, an internal model, and a sensory information processor. The behavior generator is responsible for action decomposition, turning higher level goals or missions into physical action at the human-machine interface. The internal model is an internal representation of the environment which the human is assumed to possess and is divided into four submodel categories. The sensory information processor is responsible for sensory composition. All three parts of the model act in consort to allow anticipatory behavior on the part of the human in goal-directed interaction with dynamic systems. Human workload and error are interpreted in this framework, and the familiar example of an automobile commute is used to illustrate the nature of the activity in the three model elements. Finally, with the qualitative model as a guide, verbal protocols from a manned simulation study of a helicopter instrument landing task are analyzed with particular emphasis on the effect of automation on human-machine performance.

Hess, Ronald A.↗

NASA NTRS1987

A Qualitative Model of Human Interaction with Complex Dynamic Systems

A qualitative model describing human interaction with complex dynamic systems is developed. The model is hierarchical in nature and consists of three parts: a behavior generator, an internal model, and a sensory information processor. The behavior generator is responsible for action decomposition, turning higher level goals or missions into physical action at the human-machine interface. The internal model is an internal representation of the environment which the human is assumed to possess and is divided into four submodel categories. The sensory information processor is responsible for sensory composition. All three parts of the model act in consort to allow anticipatory behavior on the part of the human in goal-directed interaction with dynamic systems. Human workload and error are interpreted in this framework, and the familiar example of an automobile commute is used to illustrate the nature of the activity in the three model elements. Finally, with the qualitative model as a guide, verbal protocols from a manned simulation study of a helicopter instrument landing task are analyzed with particular emphasis on the effect of automation on human-machine performance.

Hess, Ronald A.↗

NASA NTRSDate not supplied

Knowledge Discovery for Early Failure Assessment of Complex Engineered Systems Using Natural Language Processing

Emerging complex engineered systems may have unexpected safety issues due to novel operational environments, increasing autonomy, human-machine interaction, and other factors. To prevent failures in operation or testing that necessitate costly redesign, it is desirable to predict likely failure modes early in the design process. Information about past engineering failures in natural language format presents one possible solution by enabling the retrieval of information that can inform new designs. However, identifying documents containing usable information and extracting the required information can be prohibitively time-consuming when implemented at scale. In this research, an automated natural language processing (NLP) framework is proposed to discover relevant knowledge from documents containing failure-related design information. The framework is applied to NASA’s Lessons Learned Information System (LLIS),which is publicly available. Documents containing usable information are filtered using two different NLP-based models. Next, from the identified usable documents, a failure taxonomy is extracted using a partitioned hierarchical topic modeling approach. Partitions of the document describe different sections of the failure taxonomy – i.e., failure, cause of failure, and recommendations – as indicated by the structure of the original document. The extracted failure taxonomy can be leveraged in early design failure assessment methods. Moreover, the framework can be used to identify documents containing usable failure-related design information from other databases and extract relevant information from these documents.

Documentation and Information Science↗

NASA NTRSDate not supplied

Knowledge Discovery for Early Failure Assessment of Complex Engineered Systems Using Natural Language Processing

Emerging complex engineered systems may have unexpected safety issues due to novel operational environments, increasing autonomy, human-machine interaction, and other factors. To prevent failures in operation or testing that necessitate costly redesign, it is desirable to predict likely failure modes early in the design process. Information about past engineering failures in natural language format presents one possible solution by enabling the retrieval of information that can inform new designs. However, identifying documents containing usable information and extracting the required information can be prohibitively time-consuming when implemented at scale. In this research, an automated natural language processing (NLP) framework is proposed to discover relevant knowledge from documents containing failure-related design information. The framework is applied to NASA’s Lessons Learned Information System (LLIS),which is publicly available. Documents containing usable information are filtered using two different NLP-based models. Next, from the identified usable documents, a failure taxonomy is extracted using a partitioned hierarchical topic modeling approach. Partitions of the document describe different sections of the failure taxonomy – i.e., failure, cause of failure, and recommendations – as indicated by the structure of the original document. The extracted failure taxonomy can be leveraged in early design failure assessment methods. Moreover, the framework can be used to identify documents containing usable failure-related design information from other databases and extract relevant information from these documents.

Documentation and Information Science↗

NASA NTRS2023

Survival of Juvenile Florida Scrub-Jays is Habitat Specific, Positively Correlated With Month and Negatively Correlated With Male Breeder Death

Juvenile survival in birds is difficult to estimate but this vital rate can be an important consideration for management decisions. We estimated juvenile survival of cooperatively breeding Florida Scrub-Jays (Aphelocoma coerulescens) in a landscape degraded by fire suppression and fragmentation using data from marked (n = 325) and unmarked juveniles (n = 1,306) with an integrated hierarchical Bayesian model. To assess the combined analyses, we also analyzed these datasets separately, with a Cormack-Jolly-Seber (CJS) model (marked) and young model (unmarked). Our data consisted of monthly censuses of territorial family groups from Florida Scrub-Jay populations in East Central Florida collected over a 22-yr period. Juvenile survival was estimated from July when young Florida Scrub-Jays begin developing independence to March when they become first year individuals and grouped according to the habitat quality class of their natal territory that were based on shrub height (with intermediate shrub heights being optimal and short and tall shrub heights being suboptimal) and the presence of sandy openings (the preferred open having many sandy openings; closed not having enough). Parameter estimates in the combined analysis were intermediate to the separate analyses. Notable differences among the separate analyses were that suboptimal habitat survival was lower in the unmarked analysis, the unmarked analysis showed a linear effect of time not seen in the marked analysis, and there was an effect of male breeder death in the marked but not unmarked analysis. The combined data analysis provided more inference than did either data set analyzed separately including juveniles in optimal-closed territories unexpectedly had higher survival than those in optimal-open, survival increased through time, and male breeder death had a negative effect on survival. This study suggests that optimal-closed habitat may play an important role in juvenile Florida Scrub-Jay survival perhaps by providing better cover from predators and warrants further investigation for management implications.

Florida Scrub-Jay↗

NASA NTRS1991

Void statistics of the CfA redshift survey

Clustering properties of two samples from the CfA redshift survey, each containing about 2500 galaxies, are studied. A comparison of the velocity distributions via a K-S test reveals structure on scales comparable with the extent of the survey. The void probability function (VPF) is employed for these samples to examine the structure and to test for scaling relations in the galaxy distribution. The galaxy correlation function is calculated via moments of galaxy counts. The shape and amplitude of the correlation function roughly agree with previous determinations. The VPFs for distance-limited samples of the CfA survey do not match the scaling relation predicted by the hierarchical clustering models. On scales not greater than 10/h Mpc, the VPFs for these samples roughly follow the hierarchical pattern. A variant of the VPF which uses nearly all the data in magnitude-limited samples is introduced; it accounts for the variation of the sampling density with velocity in a magnitude-limited survey.

Vogeley, Michael S.↗

NASA NTRS2019

Bridging the Gap Between Requirements and Model Analysis: Evaluation on Cyber-Physical Challenge Problems

We present a framework for introducing high-level requirement specifications in the automated analysis of dataflow models. By integrating the FRET requirements elicitation tool with the CoCoSim analyzer, our framework enables the analysis of hierarchical Simulink models against requirements written in a restricted English language. More precisely, we support: automatic extraction of Simulink model information and association of high-level requirements with target model signals and components; translation of temporal logic formulas into synchronous dataflow CoCoSpec specifications as well as Simulink monitors; and interpretation of counterexamples produced by the analysis both at the requirement and model level. The features provided by our framework are generic and can be used to integrate other requirements elicitation and Simulink/Lustre analysis tools. We report on the lessons learned from the application of our approach to the Lockheed Martin Cyber-Physical, aerospace-inspired challenge problems. For the analysis, we used the Kind2, Zustre, and Simulink Design Verifier (SLDV) tools.

Simulink↗

DOE OSTI2024

Macroscopic modeling of gas permeability in hierarchical micro/nanoporous media: A unified characterization of rarefaction using Klinkenberg theory and equivalent diameter

Estimating gas transport through a hierarchical micro/nanoporous system is challenging due to non-equilibrium gas dynamics. The primary difficulty lies in determining the rarefaction level, because identifying a representative flow dimension in a complex porous system with multiple pore scales is not straightforward. Our study performed a pore-level analysis for gas permeability in dual-scale porous media with varying porosity, throat size, and secondary pore size under different rarefaction conditions. We found that secondary porosity negatively affects permeability due to increased friction forces, with this influence growing as the secondary pore size and porosity increase until the secondary pore becomes comparable to the throat. However, rarefaction reduces the effects of secondary pores due to boundary slip. Traditional Knudsen number (Kn) calculations based on Darcy-defined height failed to accurately describe the rarefaction effects on gas permeability. Instead, we introduced an equivalent diameter to calculate the Kn, which provided an accurate normalization of apparent gas permeability independent of pore geometry. Furthermore, the extended Kozeny–Carman–Klinkenberg model developed in our previous study successfully yielded a macroscopic model for apparent gas permeability in hierarchical micro/nanoporous systems as a function of the traditional Darcy height and porosity.

04 OIL SHALES AND TAR SANDS↗

NASA NTRS2015

Mechanical Properties of Graphene Nanoplatelet/Carbon Fiber/Epoxy Hybrid Composites: Multiscale Modeling and Experiments

Because of the relatively high specific mechanical properties of carbon fiber/epoxy composite materials, they are often used as structural components in aerospace applications. Graphene nanoplatelets (GNPs) can be added to the epoxy matrix to improve the overall mechanical properties of the composite. The resulting GNP/carbon fiber/epoxy hybrid composites have been studied using multiscale modeling to determine the influence of GNP volume fraction, epoxy crosslink density, and GNP dispersion on the mechanical performance. The hierarchical multiscale modeling approach developed herein includes Molecular Dynamics (MD) and micromechanical modeling, and it is validated with experimental testing of the same hybrid composite material system. The results indicate that the multiscale modeling approach is accurate and provides physical insight into the composite mechanical behavior. Also, the results quantify the substantial impact of GNP volume fraction and dispersion on the transverse mechanical properties of the hybrid composite, while the effect on the axial properties is shown to be insignificant.

molecular dynamics↗

NASA NTRS2015

Mechanical Properties of Graphene Nanoplatelet Carbon Fiber Epoxy Hybrid Composites: Multiscale Modeling and Experiments

Because of the relatively high specific mechanical properties of carbon fiber/epoxy composite materials, they are often used as structural components in aerospace applications. Graphene nanoplatelets (GNPs) can be added to the epoxy matrix to improve the overall mechanical properties of the composite. The resulting GNP/carbon fiber/epoxy hybrid composites have been studied using multiscale modeling to determine the influence of GNP volume fraction, epoxy crosslink density, and GNP dispersion on the mechanical performance. The hierarchical multiscale modeling approach developed herein includes Molecular Dynamics (MD) and micromechanical modeling, and it is validated with experimental testing of the same hybrid composite material system. The results indicate that the multiscale modeling approach is accurate and provides physical insight into the composite mechanical behavior. Also, the results quantify the substantial impact of GNP volume fraction and dispersion on the transverse mechanical properties of the hybrid composite, while the effect on the axial properties is shown to be insignificant.

computational mechanics↗

NASA NTRS2015

Mechanical Properties of Graphene Nanoplatelet/Carbon Fiber/Epoxy Hybrid Composites: Multiscale Modeling and Experiments

Because of the relatively high specific mechanical properties of carbon fiber/epoxy composite materials, they are often used as structural components in aerospace applications. Graphene nanoplatelets (GNPs) can be added to the epoxy matrix to improve the overall mechanical properties of the composite. The resulting GNP/carbon fiber/epoxy hybrid composites have been studied using multiscale modeling to determine the influence of GNP volume fraction, epoxy crosslink density, and GNP dispersion on the mechanical performance. The hierarchical multiscale modeling approach developed herein includes Molecular Dynamics (MD) and micromechanical modeling, and it is validated with experimental testing of the same hybrid composite material system. The results indicate that the multiscale modeling approach is accurate and provides physical insight into the composite mechanical behavior. Also, the results quantify the substantial impact of GNP volume fraction and dispersion on the transverse mechanical properties of the hybrid composite while the effect on the axial properties is shown to be insignificant.

nanocomposites↗

DOE OSTI2021

A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media

This paper presents a new deep learning data-driven model for predicting structure dependent pore-fluid velocity fields in rock. The model is based on a Convolutional Auto-Encoder (CAE) artificial neural network capable of learning from image data generated by direct numerical simulations of fluid flow through pore-structures, such as by Lattice Boltzmann or molecular dynamics methods. The main novelty of the model in comparison to previous CAE-based data-driven approaches consists of three parts. The first is a methodology for decomposing the full-domain of the porous media into sub-regions, or “sub-domains”, in order to reduce the overall size of the CAE, batch process the sub-domains in parallel, and enable the CAE to learn local and generalizable nonlinear mappings of pore-fluid velocities. The second consists of embedding the finite difference solutions of the incompressible Navier-Stokes and continuity equations into convolutional layers prior to the CAE in order to provide the CAE with knowledge of fluid dynamics physics (PhyFlow). The third main novelty is that the training of the CAE is regularized with a hierarchical loss function that encourages the learning of fluid flow patterns (in a way similar to ranked modes in principal component analysis), ranking from most to least important. This is shown to increase the stability in learning, reduce over-fitting, and promote interpretability of the CAE neural network layers (HierCAE). The comprehensive new data-driven model, which we call the PhyFlow-HierCAE model, is shown to exhibit improved accuracy and generalizability of flow field predictions over conventional CAE models, attributable to the embedded physical knowledge and the hierarchical regularization, as well as realize orders of magnitude speed-ups in computation times as a surrogate for the direct numerical simulations. Examples of training and forward predictions on unseen pore-structures are provided and evaluated for data from Lattice Boltzmann and molecular dynamics simulations of pore-fluid flow. The model is shown to be a fast and accurate emulator (or “surrogate”) for predicting effective permeability of unseen pore-structures based on learning from relatively small direct numerical simulation datasets.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2026

Semi-Analytical Hierarchical Bayesian Inference of Nonlinear Model Structure in Stochastic Dynamics: Applied to Compartmental Models of Infectious Diseases

A Bayesian computational framework for parsimonious inference in stochastic nonlinear dynamical systems is presented. This framework enables the concurrent estimation of system states, time-varying parameters, time-invariant parameters, and the optimal sparsity structure of the model parameters. Because differential equation-based models are often simplified mechanistic or phenomenological representations, robust inference from noisy measurement data requires explicit treatment of model error and uncertainty. Model error and time-varying parameters can be represented as random processes, enabling inference while making minimal assumptions about the underlying sources of discrepancy and variability. Adopting stochastic differential equation representations affords the model significant flexibility, but can also render it susceptible to overfitting during statistical inversion, where the inferred model may track noise rather than the underlying signal. To alleviate the effects of overfitting and to enable the discovery of the optimal sparse representation of the time-invariant parameters, a Bayesian sparse learning algorithm is embedded within the framework. This sparse learning framework adopts an approximate hierarchical Bayesian setting defined by a series of semi-analytical expressions. The model structure inference framework is validated using a stochastic compartmental model for tracking and forecasting active cases of an infectious disease. Compartmental models describe population-level infectious disease dynamics through interactions among population fractions grouped by disease state. Mathematically, such models consist of a system of coupled ordinary differential equations. This example adopts an expressive compartmental model that includes multiple possible interactions between disease states, motivated by early uncertainty surrounding COVID-19 reinfection dynamics and their implications for long-term epidemic forecasting. The sparse learning exercise permits the inference of a priori unknown epidemiological dynamics from simulated public health data, discovering the nested compartmental model that optimizes the trade-off between average data-fit and model complexity. It is shown that inducing sparsity among the model parameters eliminates redundant interactions between compartments, equivalently revealing the optimal coupling structure between differential equations.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2023

Experimental Observations of the Topology of Convolutional Neural Network Activations

Topological data analysis (TDA) is a branch of computational mathematics, bridging algebraic topology and data science, that provides compact, noise-robust representations of complex structures. Deep neural networks (DNNs) learn millions of parameters associated with a series of transformations defined by the model architecture resulting in high-dimensional, difficult to interpret internal representations of input data. As DNNs become more ubiquitous across multiple sectors of our society, there is increasing recognition that mathematical methods are needed to aid analysts, researchers, and practitioners in understanding and interpreting how these models' internal representations relate to the final classification. In this paper we apply cutting edge techniques from TDA with the goal of gaining insight towards interpretability of convolutional neural networks used for image classification. We use two common TDA approaches to explore several methods for modeling hidden layer activations as high-dimensional point clouds, and provide experimental evidence that these point clouds capture valuable structural information about the model's process. First, we demonstrate that a distance metric based on persistent homology can be used to quantify meaningful differences between layers and discuss these distances in the broader context of existing representational similarity metrics for neural network interpretability. Second, we show that a mapper graph can provide semantic insight as to how these models organize hierarchical class knowledge at each layer. These observations demonstrate that TDA is a useful tool to help deep learning practitioners unlock the hidden structures of their models.

topological data analysis, deep learning↗

NASA NTRS2005

Lessons Learned in the Livingstone 2 on Earth Observing One Flight Experiment

The Livingstone 2 (L2) model-based diagnosis software is a reusable diagnostic tool for monitoring complex systems. In 2004, L2 was integrated with the JPL Autonomous Sciencecraft Experiment (ASE) and deployed on-board Goddard's Earth Observing One (EO-1) remote sensing satellite, to monitor and diagnose the EO-1 space science instruments and imaging sequence. This paper reports on lessons learned from this flight experiment. The goals for this experiment, including validation of minimum success criteria and of a series of diagnostic scenarios, have all been successfully net. Long-term operations in space are on-going, as a test of the maturity of the system, with L2 performance remaining flawless. L2 has demonstrated the ability to track the state of the system during nominal operations, detect simulated abnormalities in operations and isolate failures to their root cause fault. Specific advances demonstrated include diagnosis of ambiguity groups rather than a single fault candidate; hypothesis revision given new sensor evidence about the state of the system; and the capability to check for faults in a dynamic system without having to wait until the system is quiescent. The major benefits of this advanced health management technology are to increase mission duration and reliability through intelligent fault protection, and robust autonomous operations with reduced dependency on supervisory operations from Earth. The work-load for operators will be reduced by telemetry of processed state-of-health information rather than raw data. The long-term vision is that of making diagnosis available to the onboard planner or executive, allowing autonomy software to re-plan in order to work around known component failures. For a system that is expected to evolve substantially over its lifetime, as for the International Space Station, the model-based approach has definite advantages over rule-based expert systems and limit-checking fault protection systems, as these do not scale well. The model-based approach facilitates reuse of the L2 diagnostic software; only the model of the system to be diagnosed and telemetry monitoring software has to be rebuilt for a new system or expanded for a growing system. The hierarchical L2 model supports modularity and expendability, and as such is suitable solution for integrated system health management as envisioned for systems-of-systems.

Hayden, Sandra C.↗