Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “network 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 1,405 records · Page 78
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Identifying Bayesian optimal experiments for uncertain biochemical pathway models

Abstract Pharmacodynamic (PD) models are mathematical models of cellular reaction networks that include drug mechanisms of action. These models are useful for studying predictive therapeutic outcomes of novel drug therapies in silico. However, PD models are known to possess significant uncertainty with respect to constituent parameter data, leading to uncertainty in the model predictions. Furthermore, experimental data to calibrate these models is often limited or unavailable for novel pathways. In this study, we present a Bayesian optimal experimental design approach for improving PD model prediction accuracy. We then apply our method using simulated experimental data to account for uncertainty in hypothetical laboratory measurements. This leads to a probabilistic prediction of drug performance and a quantitative measure of which prospective laboratory experiment will optimally reduce prediction uncertainty in the PD model. The methods proposed here provide a way forward for uncertainty quantification and guided experimental design for models of novel biological pathways.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

CSB-RNN: A Faster-Than-Realtime RNN Acceleration Framework with Compressed Structured Blocks

Recurrent Neural Networks (RNN) is widely applied to temporal sequence analysis, where real-time performance is usually in demand. However, RNN suffers a heavy computational workload as the model comes with a large weight matrix. To alleviate the pain, model compression (pruning) schemes have been proposed for RNN that pruning the redundant (near-zero) weight-values. On the one hand, the non-structured pruning methods achieve a considerable pruning rate while bringing the computational irregularity, which is un-friendly to parallel-hardware. On the other hand, the existing structured pruning methods consider the hardware parallelism; However, they suffer a poor pruning rate due to the restrict constraints on pruning structure. This paper presents CSB-RNN, an optimized full-stack RNN framework with the novel compressed structured block (CSB) technique. The CSB-pruned RNN model comes with both fine-granularity that benefits the pruning rate and regular structure that facilitates the hardware-parallelism. Further, we propose a novel hardware architecture for inferencing the CSB-pruned model. Different from conventional parallel hardware, this architecture solves the block-workload imbalance issue and achieves an over 95% hardware utilization. With the experiments on 10 RNN models in 5 application domains, the CSB-RNN realizes 7×-20× lossless compression and up to 50× acceptable lossy-compression, which is 2×-7× to the prior art. With the addition of the novel hardware, the compressed-RNN inference reaches a super real-time latency of 10-400µs with FPGA implementation.

Shi, Runbin↗
DOE OSTI2025

Quantum optimal control of superconducting qubits based on machine-learning characterization

Implementing fast and high-fidelity quantum operations using open-loop quantum optimal control relies on having an accurate model of the quantum dynamics. Any deviations between this model and the complete dynamics of the device, such as the presence of spurious modes or pulse distortions, can degrade the performance of optimal controls in practice. Here, we propose an experimentally simple approach to realize optimal quantum controls tailored to the device parameters and environment while specifically characterizing this quantum system. Concretely, we use physics-inspired machine learning to infer an accurate model of the dynamics from experimentally available data and then optimize our experimental controls on this trained model. We show the power and feasibility of this approach by optimizing arbitrary single-qubit operations in detailed numerical simulations of a superconducting transmon qubit. Furthermore, we demonstrate that this framework produces an accurate description of the device dynamics under arbitrary controls, together with the precise pulses achieving arbitrary single-qubit gates with a high fidelity of ∼99.99%.

Artificial neural networks↗
DOE OSTI2023

On the influence of over-parameterization in manifold based surrogates and deep neural operators

Constructing accurate and generalizable approximators (surrogate models) for complex physico-chemical processes exhibiting highly non-smooth dynamics is challenging. The main question is what type of surrogate models we should construct and should these models be under-parameterized or over-parameterized. In this work, we propose new developments and perform comparisons for two promising approaches: manifold-based polynomial chaos expansion (m-PCE) and the deep neural operator (DeepONet), and we examine the effect of over-parameterization on generalization. While m-PCE enables the construction of a mapping by first identifying low-dimensional embeddings of the input functions, parameters, and quantities of interest (QoIs), a neural operator learns the nonlinear mapping via the use of deep neural networks. Here, we demonstrate the performance of these methods in terms of generalization accuracy by solving the 2D time-dependent Brusselator reaction-diffusion system with uncertainty sources, modeling an autocatalytic chemical reaction between two species. We first propose an extension of the m-PCE by constructing a mapping between latent spaces formed by two separate embeddings of the input functions and the output QoIs. To further enhance the accuracy of the DeepONet, we introduce weight self-adaptivity in the loss function. We demonstrate that the performance of m-PCE and DeepONet is comparable for cases of relatively smooth input-output mappings. However, when highly non-smooth dynamics is considered, DeepONet shows higher approximation accuracy. We also find that for m-PCE, modest over-parameterization leads to better generalization, both within and outside of distribution, whereas aggressive over-parameterization leads to over-fitting. In contrast, an even highly over-parameterized DeepONet leads to better generalization for both smooth and non-smooth dynamics. Furthermore, we compare the performance of the above models with another recently proposed operator learning model, the Fourier Neural Operator, and show that its over-parameterization also leads to better generalization. Taken together, our studies show that m-PCE can provide very good accuracy at very low training cost, whereas a highly over-parameterized DeepONet can provide better accuracy and robustness to noise but at higher training cost. In both methods, the inference cost is negligible.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Machine Learning Models for Binary Molecular Classification using VUV Absorption Spectra

Machine learning methods were combined with differential absorption spectroscopy measurements in the vacuum-ultraviolet region (5.167 – 9.920 eV) in order to develop predictive capabilities for inferring molecular structure from the spectra. Several types of species were analyzed and, for modeling purposes, were defined using a single classification: (1) alkane, (2) conjugation with oxygen (e.g. diacetyl, ethyl vinyl ether), (3) non-conjugated alkene (e.g. 1-butene, 1,4-cyclohexadiene), (4) oxygen-containing (e.g. 1-butanol, tetrahydrofuran), or (5) cyclic (e.g. cyclopentane, cyclohexanone). The latter molecular classification excluded cyclic ethers. Several modeling methods were employed in the analysis of 102 absorption spectra, 24 of which were measured for the first time. The primary objective was to identify suitable methods that enable accurate predictions of molecular structure classifications with minimized statistical uncertainties. Rather than identifying a single, unifying method to reliably predict molecular structure contributions to VUV absorption spectra, coordination is required among a particular method, the type of molecular structure detail (e.g. conjugation), and absorption region of interest. The latter is accomplished using a binning approach, wherein absorption regions of ~0.5 eV were utilized rather than the entire ~4.8 eV range. Photon energy binning enabled analysis of region-specific predictions of accuracy, precision, and recall. The outcome from the binning approach is that, rather than utilizing the entire spectrum, optimal determination of molecular structure using machine learning methods depends on the absorption region. Furthermore, the present work provides separate machine learning models for each molecular classification, which enables the identification of multi-functional species relevant to atmospheric chemistry and combustion chemistry, where isomer-resolved speciation is critical to understanding complex reaction networks.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Comparative Performance Evaluation of Large Language Models for Extracting Molecular Interactions and Pathway Knowledge

Understanding the interactions and regulatory relationships among biomolecules is essential for deciphering complex biological systems and elucidating the mechanisms behind diverse biological functions. Traditionally, the collection of such molecular interaction data has relied on expert curation, a process that is both time-consuming and labor-intensive. To address these limitations, this study explores the use of large language models (LLMs) to automate the genome-scale extraction of molecular interaction knowledge. Here, we evaluate the performance of various LLMs on key biological tasks, including the identification of protein-protein interactions, detection of genes associated with pathways influenced by low-dose radiation, and inference of gene regulatory relationships. Our findings demonstrate that larger LLMs tend to perform better, particularly in extracting intricate gene and protein interactions. Despite their strengths, these models face challenges in recognizing functionally diverse gene groups and highly correlated regulatory relationships. Through a comprehensive analysis using established molecular interaction and pathway databases, we show that LLMs possess the potential to identify relevant biomolecules and predict their interactions, offering valuable insights and marking a significant step toward AI-driven biological knowledge discovery.

63 RADIATION, THERMAL, AND OTHER ENVIRON. POLLUTAN↗
DOE OSTI2025

Graph reinforcement learning for exploring model spaces beyond the standard model

We present a methodology for performing scans of beyond the standard model (BSM) parameter spaces with reinforcement learning. We identify a novel procedure using graph neural networks that is capable of exploring spaces of models without the user specifying a fixed particle content, allowing broad classes of BSM models to be explored—in theory, the technique is applicable to nearly any model space with a prespecified gauge group. We provide a generic procedure by which a suitable graph grammar can be developed for any BSM model that features user-specified symmetry groups and a finite number of different possible particle species, the use of which is applicable to a variety of machine learning tasks over the actions of BSM theories beyond our particular reinforcement learning use case. As a proof of concept, we construct the graph grammar for theories with vectorlike leptons that may or may not be charged under a dark U ( 1 ) group, inspired by portal matter extensions of the sub-GeV vector portal/kinetic mixing simplified dark matter models. We then use this graph grammar to create a reinforcement learning environment tasked with creating models with these vectorlike leptons that are consistent with a list of a variety of precision observables. The reinforcement learning agent succeeds in developing models that can address the observed muon anomalous magnetic moment discrepancy while remaining consistent with flavor violation and electroweak precision observables, including both constructions that have previously been studied as well as new models that have not, to our knowledge, previously been identified. By inspecting the resulting ensembles of models that the agent produces and experimenting with different configurations for our reinforcement learning environment and graph grammar, we also infer various lessons about the development of these environments that can be transferable to reinforcement learning scans of more complicated model spaces and comment on future directions for the development of this technique into a more mature tool. Published by the American Physical Society 2025

Wojcik, George N.↗
DOE OSTI2021

Physics guided machine learning using simplified theories

Recent applications of machine learning, in particular deep learning, motivate the need to address the generalizability of the statistical inference approaches in physical sciences. In this Letter, we introduce a modular physics guided machine learning framework to improve the accuracy of such data-driven predictive engines. The chief idea in our approach is to augment the knowledge of the simplified theories with the underlying learning process. To emphasize their physical importance, our architecture consists of adding certain features at intermediate layers rather than in the input layer. To demonstrate our approach, we select a canonical airfoil aerodynamic problem with the enhancement of the potential flow theory. We include the features obtained by a panel method that can be computed efficiently for an unseen configuration in our training procedure. By addressing the generalizability concerns, our results suggest that the proposed feature enhancement approach can be effectively used in many scientific machine learning applications, especially for the systems where we can use a theoretical, empirical, or simplified model to guide the learning module.

42 ENGINEERING↗
DOE OSTI2019

Multi-Formalism Modeling for Disaster Resilience, Forecasting, and Response

Disaster response preparedness, risk mitigation, and effective post-event response are critical for national security, public health, and economic integrity. For areas impacted by extreme events, timely and accurate impact assessments of lifeline infrastructures are critical. Traditional disaster management and situational awareness platforms are not capable of handling the massive influx of available information from traditional and non-traditional sources. Next-generation analytic assessment approaches, where use of dynamically sourced, heterogeneous data with near real-time availability are integrated into a continually-updated analytics framework. The objective of this research was to develop a foundational multi-formalism modeling platform to integrate novel data sources for use in improving situational awareness after a disaster event. Specifically, to 1) provide an architecture for machine-learning based data fusion various multi-sensor datasets; 2) classify, geolocate, and temporally bin novel data sources for use in spatially-enabled machine-learning based data fusion; and 3) derive confidence-based observation data with machine-learning based prediction to fill spatial and temporal observation gaps. This project developed a reusable cloud-based pipeline architecture for ingesting, storing, and analyzing relevant data, executing a range of deep-learning models, and visualize the information geographically. The pipeline development was informed by previous PNNL investments (StreamKit) and built upon leading-edge technologies consisting of: 1) a raw datastore; 2) a data enrichment pipeline with custom and third-party analytics; 3) enriched data storage; and 4) a custom UI to display the enriched data. Machine learning/deep learning models were selected and tested to classify social media images for relevancy and presence of flood waters. Our custom trained relevance models were run on a validation dataset and the overall accuracy of the relevance classifiers ranged from 65% to 90%. We found that convolutional neural network (CNN) models were both faster and more accurate than more traditional machine learning models. We applied the relevance classifier to remove non-relevant images and then undertook a limited training of CNN models to detect flood on ~1,800 labeled images. The testing accuracy of the flood detection models were ~88%. To improve the geographic location of both social media and oblique aerial images we tested a multi-step workflow built on a pre-trained deep-learning model. To identify similarity features from test and reference scenes, we applied the Xception and Faiss models to a set of test images with inaccurate spatial locations, as well and precisely-located reference scenes from the general area. Matching reference scenes to the test scenes with the closest similarity index had limited success in improving the precision of test scene location and additional methods have been identified for testing. Data fusion between authoritative geographic data (satellite imaging) and non-authoritative crowdsourced data is an underlying theme throughout this research and methods are tested bring these data to a common geographic form through feature matching and transformations. The automated extraction of supporting information for disaster response from oblique aerial images or ground-level social media images is accomplished through semantic image segmentation and transformation. The non-authoritative data are used to validate the authoritative data collections and processing and optionally, t+0 flood models. In addition, the common form geographic data are combined with other static landscape features to develop spatially-explicit training vectors used in deep generative statistical inference models to develop complex non-linear relationships in static and dynamic data and ultimately derive a spatially continuous and probabilistic assessment of flood damage with the event domain.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2014

New effects of a long-lived negatively charged massive particle on big bang nucleosynthesis

Primordial {sup 7}Li abundance inferred from observations of metal-poor stars is a factor of about 3 lower than the theoretical value of standard big bang nucleosynthesis (BBN) model. One of the solutions to the Li problem is {sup 7}Be destruction during the BBN epoch caused by a long-lived negatively charged massive particle, X{sup −}. The particle can bind to nuclei, and X-bound nuclei (X-nuclei) can experience new reactions. The radiative X{sup −} capture by {sup 7}Be nuclei followed by proton capture of the bound state of {sup 7}Be and X{sup −} ({sup 7}Be{sub x}) is a possible {sup 7}Be destruction reaction. Since the primordial abundance of {sup 7}Li originates mainly from {sup 7}Li produced via the electron capture of {sup 7}Be after BBN, the {sup 7}Be destruction provides a solution to the {sup 7}Li problem. We suggest a new route of {sup 7}Be{sub x} formation, that is the {sup 7}Be charge exchange at the reaction of {sup 7}Be{sup 3+} ion and X{sup −}. The formation rate depends on the ionization fraction of {sup 7}Be{sup 3+} ion, the charge exchange cross section of {sup 7}Be{sup 3+}, and the probability that excited states {sup 7}Be{sub x}* produced at the charge exchange are converted to the ground state. We find that this reaction can be equally important as or more important than ordinary radiative recombination of {sup 7}Be and X{sup −}. The effect of this new route is shown in a nuclear reaction network calculation.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Karhunen–Loève deep learning method for surrogate modeling and approximate Bayesian parameter estimation

We evaluate the performance of the Karhunen-Loève Deep Neural Network (KL-DNN) framework for surrogate modeling and approximate Bayesian parameter estimation in partial differential equation models. In the surrogate model, the Karhunen-Loève (KL) expansions are used for the dimensionality reduction of the number of unknown parameters and variables, and a deep neural network is employed to relate the reduced space of parameters to that of the state variables. The KL-DNN surrogate model is used to formulate a maximum-a-posteriori-like least-squares problem, which is randomized to draw samples of the posterior distribution of the parameters. We test the proposed framework for a hypothetical unconfined aquifer via comparison with the forward MODFLOW and inverse PEST++ iterative ensemble smoother (IES) solutions as well as the state-of-the-art Fourier neural operator (FNO) and deep operator networks (DeepONets) operator learning surrogate models. Our results show that the KL-DNN surrogate model outperforms FNO and DeepONet for forward predictions. For solving inverse problems, the randomized algorithm provides the same or more accurate Bayesian predictions of the parameters than IES as evidenced by the higher log-predictive probability of both the estimated parameter field and the forecast hydraulic head. The posterior mean obtained from the randomized algorithm is closer to the reference parameter field than that obtained with FNO as the maximum a posteriori estimate.

Approximate Bayesian inference↗
NASA NTRS2008

Total Lightning Observations of Extreme Weather Events over the Contiguous United States

The overall objective is to investigate total lightning characteristics of extreme weather events over the contiguous United States (CONUS) using TRMM (Tropical Rainfall Measuring Mission) LIS (Lightning Image Sensor) and OTD (Optical Transient Detector) satellite observations. A large LIS (10+ years) and OTD (5 years) data base is available to study the instantaneous total or cloud-to-ground (CG) plus intracloud (IC) lightning characteristics of extreme weather events. More specifically, the LIS and OTD data are combined with National Lightning Detection Network (NLDN) observations to examine the total and CG lightning flash rate and density, the IC:CG ratio, and positive CG percentage. These instantaneous lightning characteristics can be used for basic science studies to better understand the physical and dynamical linkages between lightning and precipitation and their environmental controls. They can also provide a first-look of extreme weather events leading up to future satellite observations (e.g., NOAA GOES-R Geostationary Lightning Mapper [GLM]) for use in climate studies and the short-term prediction and warning process. Extreme weather events are defined by the NOAA Storm Data reports of tornadoes, large hail (at least 0.75 inch) and strong straight-line winds (at least 50 kts). Over CONUS, there are over 70,000 severe storm reports in the TRMM spatial domain (< 35 N) from 1998-2007 and over 100,000 storm reports in the OTD spatial domain (5/1995-4/2000). Temporal co-location is on the order of 1% (i.e., 1000 s of coincident overpasses), providing a statistically significant sample of instantaneous total lightning properties. This instantaneous behavior of lightning in extreme weather is then compared to that of typical thunderstorm events, or randomly sampled LIS/OTD events in which the extreme events have been eliminated from the population. Results describing the instantaneous behavior of total lightning within a large sample of extreme and typical thunderstorms over CONUS will be presented. When possible, coincident VHF lightning observations from the ground-based Northern Alabama Lightning Mapping Array (LMA) are compared to the LIS/OTD optical lightning signatures, providing a validation source for instantaneous space-based optical lightning properties and a means to extend lightning inferences over the life-cycle of extreme weather.

Carey, Lawrence D.↗
DOE OSTI2023

Fast 2D Bicephalous Convolutional Autoencoder for Compressing 3D Time Projection Chamber Data

High-energy large-scale particle colliders produce data at high speed in the order of 1 terabytes per second in nuclear physics and petabytes per second in high energy physics. Developing real-time data compression algorithms to reduce such data at high throughput to fit permanent storage has drawn increasing attention. Specifically, at the newly constructed sPHENIX experiment at the Relativistic Heavy Ion Collider (RHIC), a time projection chamber is used as the main tracking detector, which records particle trajectories in a volume of three-dimensional (3D) cylinder. The resulting data are usually very sparse with occupancy around 10.8%. Such sparsity presents a challenge to conventional learning-free lossy compression algorithms, such as SZ, ZFP, and MGARD. The 3D convolutional neural network (CNN)-based approach, Bicephalous Convolutional Autoencoder (BCAE), outperforms traditional methods both in compression rate and reconstruction accuracy. BCAE can also utilize the computation power of graphical processing units suitable for deployment in a modern heterogeneous highperformance computing environment. This work introduces two BCAE variants: BCAE++ and BCAE-2D. BCAE++ achieves a 15% better compression ratio and a 77% better reconstruction accuracy measured in mean absolute error compared with BCAE. BCAE-2D treats the radial direction as the channel dimension of an image, resulting in a 3× speedup in compression throughput. In addition, we demonstrate an unbalanced autoencoder with a larger decoder can improve reconstruction accuracy without significantly sacrificing throughput. Lastly, we observe both the BCAE++ and BCAE-2D can benefit more from using half-precision mode in throughput (76 - 79% increase) without loss in reconstruction accuracy. The source code and links to data and pretrained models can be found at https://github.com/BNL-DAQ-LDRD/NeuralCompression_v2

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
NASA NTRS2018

Radar Measurements of the LEO Orbital Debris Environment

Access to space and the preservation of the near-Earth space environment is of critical significance. Increased interest in issues surrounding space traffic management and the continued assessment and discussion of orbital debris at the United Nations Committee on the Peaceful Uses of Outer Space (COPUOS) illustrates the significance of the topic of orbital debris. There are currently over 20,000 tracked objects in the publicly available satellite catalog on Space-Track.org. The catalog is maintained by the US Air Force Space Command using a network of optical and radar ground-based sensors and is believed to be complete for a characteristic size of 10 cm or larger in low Earth orbit (LEO). Based on the work of the NASA Orbital Debris Program Office (ODPO) over approximately the past 40 years it is understood that the small debris population (< 10 cm) increases by orders of magnitude as characteristic size decreases. With population estimates ranging from 500,000 to 1,000,000 small debris objects (> 5 mm) in orbit, it is currently not practical to track and maintain precision orbits on every object. Instead the NASA ODPO uses powerful ground-based radars to sample the low Earth Orbit (LEO) environment and assign approximate orbits to each detection. This poses an interesting signal processing challenge as we are trying to detect the smallest objects possible on the edge of the radar's sensitivity. For approximately the last 30 years, NASA ODPO has partnered with the Massachusetts Institute of Technology Lincoln Laboratory (MIT/LL) to utilize the Haystack Ultra-wideband Satellite Imaging Radar (HUSIR - formerly the Long-Range Imaging Radar or simply Haystack) and the Haystack Auxiliary (HAX) radar to collect orbital debris radar data. Additionally, the ODPO collaborates with the NASA Jet Propulsion Laboratory (JPL) to use the Goldstone Solar System Radar. The orbital debris detections from these radars serve as inputs for statistical risk models used by the human spaceflight and satellite communities to assess risk to spacecraft posed by orbital debris. In this paper, we will describe the history of orbital debris radar measurements conducted by NASA, provide an overview of current radar measurements techniques and facilities, discuss the signal processing software used for orbital debris measurements and the inference of debris size and orbital parameters from these measurements, and discuss how orbital debris radar measurements are validated for use in models that are used throughout the aerospace industry.

Blackwell, Chris↗
DOE OSTI2020

NEON AOP foliar trait maps, maps of model uncertainty estimates, and conifer map, East River, CO 2018

This data package contains mapped trait estimates and their uncertainties, and conifer map, for the National Ecological Observatory Network's Airborne Observation Platform survey data acquired over the Upper East River, Colorado in 2018. For full details, please see associated reference. in brief, trait models were developed independently for needle and non-needle leaf species using partial least squares regression (PLSR) using ground data from additional datasets: doi:10.15485/1618130, doi:10.15485/1618132, and doi:10.15485/1631278, merged with extracted spectral data from doi:10.15485/1618131. We separated vegetated pixels into needle and non-needle classes in order to generate a classification map based on the spectral differences between these leaf types (conifer.tif). We trained a deep learning model with custom architecture, detailed in Chadwick et al. In Press. The model performed with 0.998 true positive rate and 0.982 true negative rate, with ‘positives’ being non-needle identification. We then utilized PLSR to generate models of foliar traits for each leaf type. So that we could also map uncertainty in these predictions, we generated ten different models for needle and non-needle leaf species using different testing holdout sets of discrete sites. Each of these models was developed with a 100-fold cross validation procedure that utilized a 70% training set and 30% validation set with each fold, and then assessed based on the 10% of testing sites that were not included in that model’s development. The mean predicted value across the 10 models is used for the trait estimate in each pixel across the study area. The models are applied according to the leaf type designation in the conifer.tif map. The errors are the standard deviation across the 10 different models developed, with high error suggesting instability in model prediction and areas where values may not be reliable for ecological inference. These maps are only applied to areas with a NDVI > 0.5 to exclude non-vegetated areas. Shade masks could be applied to these data (doi:10.15485/1618131), but have not been for this data package. These data are also available on Google Earth Engine: https://code.earthengine.google.com/?asset=users/kdc/ER_NEON

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2013

ACCURATE CHARACTERIZATION OF HIGH-DEGREE MODES USING MDI OBSERVATIONS

We present the first accurate characterization of high-degree modes, derived using the best Michelson Doppler Imager (MDI) full-disk full-resolution data set available. A 90 day long time series of full-disk 2 arcsec pixel{sup -1} resolution Dopplergrams was acquired in 2001, thanks to the high rate telemetry provided by the Deep Space Network. These Dopplergrams were spatially decomposed using our best estimate of the image scale and the known components of MDI's image distortion. A multi-taper power spectrum estimator was used to generate power spectra for all degrees and all azimuthal orders, up to l = 1000. We used a large number of tapers to reduce the realization noise, since at high degrees the individual modes blend into ridges and thus there is no reason to preserve a high spectral resolution. These power spectra were fitted for all degrees and all azimuthal orders, between l = 100 and l = 1000, and for all the orders with substantial amplitude. This fitting generated in excess of 5.2 Multiplication-Sign 10{sup 6} individual estimates of ridge frequencies, line widths, amplitudes, and asymmetries (singlets), corresponding to some 5700 multiplets (l, n). Fitting at high degrees generates ridge characteristics, characteristics that do not correspond to the underlying mode characteristics. We used a sophisticated forward modeling to recover the best possible estimate of the underlying mode characteristics (mode frequencies, as well as line widths, amplitudes, and asymmetries). We describe in detail this modeling and its validation. The modeling has been extensively reviewed and refined, by including an iterative process to improve its input parameters to better match the observations. Also, the contribution of the leakage matrix on the accuracy of the procedure has been carefully assessed. We present the derived set of corrected mode characteristics, which includes not only frequencies, but line widths, asymmetries, and amplitudes. We present and discuss their uncertainties and the precision of the ridge-to-mode correction schemes, through a detailed assessment of the sensitivity of the model to its input set. The precision of the ridge-to-mode correction is indicative of any possible residual systematic biases in the inferred mode characteristics. In our conclusions, we address how to further improve these estimates, and the implications for other data sets, like GONG+ and HMI.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

Automatic extraction of cancer registry reportable information from free-text pathology reports using multitask convolutional neural networks

Objective: In this work we implement 2 different multitask learning (MTL) techniques, hard parameter sharing and cross-stitch, to train a word-level convolutional neural network (CNN) specifically designed for automatic extraction of cancer data from unstructured text in pathology reports. We show the importance of learning related information extraction (IE) tasks leveraging shared representations across the tasks to achieve state-of-the-art performance in classification accuracy and computational efficiency. Materials and Methods: Multitask CNN (MTCNN) attempts to tackle document information extraction by learning to extract multiple key cancer characteristics simultaneously. We trained our MTCNN to perform 5 information extraction tasks: (1) primary cancer site (65 classes), (2) laterality (4 classes), (3) behavior (3 classes), (4) histological type (63 classes), and (5) histological grade (5 classes). We evaluated the performance on a corpus of 95 231 pathology documents (71 223 unique tumors) obtained from the Louisiana Tumor Registry. We compared the performance of the MTCNN models against single-task CNN models and 2 traditional machine learning approaches, namely support vector machine (SVM) and random forest classifier (RFC). Results: MTCNNs offered superior performance across all 5 tasks in terms of classification accuracy as compared with the other machine learning models. Based on retrospective evaluation, the hard parameter sharing and cross-stitch MTCNN models correctly classified 59.04% and 57.93% of the pathology reports respectively across all 5 tasks. The baseline models achieved 53.68% (CNN), 46.37% (RFC), and 36.75% (SVM). Based on prospective evaluation, the percentages of correctly classified cases across the 5 tasks were 60.11% (hard parameter sharing), 58.13% (cross-stitch), 51.30% (single-task CNN), 42.07% (RFC), and 35.16% (SVM). Moreover, hard parameter sharing MTCNNs outperformed the other models in computational efficiency by using about the same number of trainable parameters as a single-task CNN. Conclusions: The hard parameter sharing MTCNN offers superior classification accuracy for automated coding support of pathology documents across a wide range of cancers and multiple information extraction tasks while maintaining similar training and inference time as those of a single task–specific model.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
NASA NTRS2007

Systems Modeling to Implement Integrated System Health Management Capability

ISHM capability includes: detection of anomalies, diagnosis of causes of anomalies, prediction of future anomalies, and user interfaces that enable integrated awareness (past, present, and future) by users. This is achieved by focused management of data, information and knowledge (DIaK) that will likely be distributed across networks. Management of DIaK implies storage, sharing (timely availability), maintaining, evolving, and processing. Processing of DIaK encapsulates strategies, methodologies, algorithms, etc. focused on achieving high ISHM Functional Capability Level (FCL). High FCL means a high degree of success in detecting anomalies, diagnosing causes, predicting future anomalies, and enabling health integrated awareness by the user. A model that enables ISHM capability, and hence, DIaK management, is denominated the ISHM Model of the System (IMS). We describe aspects of the IMS that focus on processing of DIaK. Strategies, methodologies, and algorithms require proper context. We describe an approach to define and use contexts, implementation in an object-oriented software environment (G2), and validation using actual test data from a methane thruster test program at NASA SSC. Context is linked to existence of relationships among elements of a system. For example, the context to use a strategy to detect leak is to identify closed subsystems (e.g. bounded by closed valves and by tanks) that include pressure sensors, and check if the pressure is changing. We call these subsystems Pressurizable Subsystems. If pressure changes are detected, then all members of the closed subsystem become suspect of leakage. In this case, the context is defined by identifying a subsystem that is suitable for applying a strategy. Contexts are defined in many ways. Often, a context is defined by relationships of function (e.g. liquid flow, maintaining pressure, etc.), form (e.g. part of the same component, connected to other components, etc.), or space (e.g. physically close, touching the same common element, etc.). The context might be defined dynamically (if conditions for the context appear and disappear dynamically) or statically. Although this approach is akin to case-based reasoning, we are implementing it using a software environment that embodies tools to define and manage relationships (of any nature) among objects in a very intuitive manner. Context for higher level inferences (that use detected anomalies or events), primarily for diagnosis and prognosis, are related to causal relationships. This is useful to develop root-cause analysis trees showing an event linked to its possible causes and effects. The innovation pertaining to RCA trees encompasses use of previously defined subsystems as well as individual elements in the tree. This approach allows more powerful implementations of RCA capability in object-oriented environments. For example, if a pressurizable subsystem is leaking, its root-cause representation within an RCA tree will show that the cause is that all elements of that subsystem are suspect of leak. Such a tree would apply to all instances of leak-events detected and all elements in all pressurizable subsystems in the system. Example subsystems in our environment to build IMS include: Pressurizable Subsystem, Fluid-Fill Subsystem, Flow-Thru-Valve Subsystem, and Fluid Supply Subsystem. The software environment for IMS is designed to potentially allow definition of any relationship suitable to create a context to achieve ISHM capability.

Figueroa, Jorge F.↗