Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Labeled Datasets”

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 469 records · Page 26

PHASE: Personalized Head-based Automatic Simulation for Electromagnetic properties in 7T MRI

Accurate and individualized human head models are becoming increasingly important for electromagnetic (EM) simulations. These simulations depend on precise anatomical representations to realistically model electric and magnetic field distributions, particularly when evaluating Specific Absorption Rate (SAR) within safety guidelines. State of the art simulations use the Virtual Population due to limited public resources and the impracticality of manually annotating patient data at scale. Here, this paper introduces Personalized Head-based Automatic Simulation for EM properties (PHASE), an automated open-source toolbox that generates high-resolution, patient-specific head models for EM simulations using paired T1-weighted (T1w) magnetic resonance imaging (MRI) and computed tomography (CT) scans with 14 tissue labels. To evaluate the performance of PHASE models, we conduct semi-automated segmentation and EM simulations on 15 real human patients, serving as the gold standard reference. The PHASE model achieved comparable global SAR and localized SAR averaged over 10 grams of tissue (SAR-10g), demonstrating its potential as a promising tool for generating large-scale human model datasets in the future. The code and models of PHASE toolbox have been made publicly available: https://github.com/hrlblab/PHASE.

Deep learning↗

Processes, data structures, and apparatuses for representing knowledge

Processes, data structures, and apparatuses to represent knowledge are disclosed. The processes can comprise labeling elements in a knowledge signature according to concepts in an ontology and populating the elements with confidence values. The data structures can comprise knowledge signatures stored on computer-readable media. The knowledge signatures comprise a matrix structure having elements labeled according to concepts in an ontology, wherein the value of the element represents a confidence that the concept is present in an information space. The apparatus can comprise a knowledge representation unit having at least one ontology stored on a computer-readable medium, at least one data-receiving device, and a processor configured to generate knowledge signatures by comparing datasets obtained by the data-receiving devices to the ontologies.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Data structures and apparatuses for representing knowledge

Data structures and apparatuses to represent knowledge are disclosed. The processes can comprise labeling elements in a knowledge signature according to concepts in an ontology and populating the elements with confidence values. The data structures can comprise knowledge signatures stored on computer-readable media. The knowledge signatures comprise a matrix structure having elements labeled according to concepts in an ontology, wherein the value of the element represents a confidence that the concept is present in an information space. The apparatus can comprise a knowledge representation unit having at least one ontology stored on a computer-readable medium, at least one data-receiving device, and a processor configured to generate knowledge signatures by comparing datasets obtained by the data-receiving devices to the ontologies.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Frost Tables, NGEE Areas B, C and D, Utqiagvik (Barrow), Alaska, 2012-2014

This dataset represents spatially intensive thaw depth surveys with individual point measurements spaced ~0.5 m apart. The three ~10x10m grids cover an ice wedge and a portion of its two neighboring polygons. The data contain thaw depth, frost table elevation, ground surface elevation, active layer depth and surface water inundation across three seasons (2012, 2013 and 2014) at Utqiagvik (Barrow) NGEE Areas B, C and D. There are a total of 69 extracted files separated by month, year, and area. The data were originally provided in excel workbooks, these files have been preserved within the zip "Original_Files." The files were transformed for archiving with images pulled out into JPEGs, data preserved as CSV files and some methodological information preserved as TXT files. Some functions, features, and formatting were lost during the transformation. Missing data were defined with blank cells. Additional documentation can be found in files labelled with "information.txt" file endings.The Next-Generation Ecosystem Experiments: Arctic (NGEE Arctic), was a research effort to reduce uncertainty in Earth System Models by developing a predictive understanding of carbon-rich Arctic ecosystems and feedbacks to climate. NGEE Arctic was supported by the Department of Energy's Office of Biological and Environmental Research.The NGEE Arctic project had two field research sites: 1) located within the Arctic polygonal tundra coastal region on the Barrow Environmental Observatory (BEO) and the North Slope near Utqiagvik (Barrow), Alaska and 2) multiple areas on the discontinuous permafrost region of the Seward Peninsula north of Nome, Alaska.Through observations, experiments, and synthesis with existing datasets, NGEE Arctic provided an enhanced knowledge base for multi-scale modeling and contributed to improved process representation at global pan-Arctic scales within the Department of Energy's Earth system Model (the Energy Exascale Earth System Model, or E3SM), and specifically within the E3SM Land Model component (ELM).[2024-07-15] Renamed BEO_Thaw_Depth_Frost_Table_Area_*_*_*.csv, BEO_Thaw_Depth_Frost_Table_Area_*_*_*_Plot_1.JPEG, and BEO_Thaw_Depth_Frost_Table_Area_*_*_*_Plot_2.JPEG names that had E or L to be Early or Late. Renamed BEO_Thaw_Depth_Frost_Table_Area_*_*_*_Plot_1.JPEG, BEO_Thaw_Depth_Frost_Table_Area_*_*_*_Plot_2.JPEG, and BEO_Thaw_Depth_Frost_Table_Area_*_information_Plot_1.JPEG to BEO_Thaw_Depth_Frost_Table_Area_*_*_*_Gridpoints.JPEG, BEO_Thaw_Depth_Frost_Table_Area_*_*_*_3Dgraphic.JPEG, and BEO_Thaw_Depth_Frost_Table_Area_*_information_LiDAR_DEM.JPEG, respectively. Updated BEO_Thaw_Depth_Frost_Table_flmd.csv file to reflect above updates.

54 ENVIRONMENTAL SCIENCES↗

HydraGNN_Predictive_GFM_2024 - Ensemble of predictive graph foundation models for ground state atomistic materials modeling

We provide the ensemble of fifteen pre-trained graph foundation models (GFMs) for atomistic materials modeling applications. Each one of the fifteen GFMs has been trained on five open-source datasets that (once aggregated) amount to over 154 million atomistic structures, which cover over two-thirds of the natural elements of the periodic table and that comprises a broad set of organic and inorganic compounds. This vast set of atomistic structures comprises ground state configurations that are dynamically stable (i.e., equilibrated structures with atomic forces approximately close to zero values) as well as dynamically unstable structures (i.e., non-equilibrium structures with non-negligible non-zero values of atomic forces). The ensemble of datasets aggregated does NOT include excited states. The datasets have been curated to remove atomistic structures with spectral norm of the force tensor above 100 eV/angstrom. Moreover, a linear term of the energy was computed for each dataset using a linear regression model that uses the chemical concentration of each natural element as regressor. The linear term predicted by the linear regression model has been subtracted from each original energy value to perform a re-alignment of the energy values across different electronic structures approximation theories performed to generate the diverse multi-source, multi-fidelity datasets. The folder "ADIOS_files" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "ADIOS_files" directory contains 6 sub-directories named as follows: - ANI1x-v3.bp - MPTrj-v3.bp - OC2020-20M-v3.bp - OC2020-v3.bp - OC2022-v3.bp - qm7x-v3.bp Each sub-directory contains the pre-processed datasets converted in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used to the development, training, and performance testing of the ensemble go predictive graph foundation models. Each GFM was developed using HydraGNN (https://github.com/ORNL/HydraGNN) as underlying graph neural network (GNN) architecture. The multi-task learning (MTL) capability of HydraGNN was used to simultaneously train the GFMs on labeled values for direct predictions of energy (a total system property of an atomistic structure that measures the chemical stability) and atomic forces (an atomic level property of an atomistic structure that measures the dynamical stability). The hyper parameters of the GFM have been tuned using scalable hyperparameter optimization (HPO) algorithms implemented in the software DeepHyper (https://github.com/deephyper/deephyper). The pre-training of each HPO trial was performed using distributed data parallelism (DDP) to scale the training across 128 compute nodes of the exascale OLCF supercomputer Frontier. Each HPO trial was trained only for 10 epochs and an early stopping was performed to avoid wasting significant computational resources on GNN architectures that were clearly underperforming. For each HPO trial, the 'omnistat' tool developed by (AMD Research - Advanced Micro Device) was used to measure the total energy consumption in kWh. The ensemble of GFMs was obtained by selecting the fifteen best performing HPO trials. Four models have been selected for their clear advantage in accuracy, and these are the GFMs with IDs 229, 156, 147, 260. Additional eleven models have been selected based on judicious balance between accuracy and energy consumption needed for training, and these are the GFMs with IDs 165, 78, 137, 1, 175, 171, 181, 67, 179, 167, 351. Each selected GFM of the ensemble was continued to cumulate a total of at most 30 epochs. In some cases, the total number of epochs actually performed was les than 30 due to two combined factors: (1) the size of the GFM (i.e., the number of model parameters to train) and (2) the total wall-clock time for which the computational resources could be allocated on OLCF-Frontier. The "Ensemble_of_models" directory contains 15 sub-directories named as follows: - gfm_0.229 - gfm_0.156 - gfm_0.147 - gfm_0.260 - gfm_0.165 - gfm_0.78 - gfm_0.137 - gfm_0.1 - gfm_0.175 - gfm_0.171 - gfm_0.181 - gfm_0.67 - gfm_0.179 - gfm_0.167 - gfm_0.351 Each one of these sub-directories refers to one of the fifteen HPO trials that have been selected to continue the pre-training with at most 30 epochs. With each sub-directory associated with a specific HPO trial, the following files can be found: - config.json: file for argument parsing to develop and train an HydraGNN architecture - gfm_0.ID_epoch_N.pk: file with model parameters for HPO ID trial after N epochs of training The ensemble of fifteen GFM architectures was used for (1) ensemble averaging to stabilize the predictions of energy and atomic forces after pre-training for post-processing analysis and (2) ensemble uncertainty quantification (UQ). The code used to develop, pre-train, and load the pre-trained models for post-processing analysis is available on the ORNL-GitHub at the following link: https://github.com/ORNL/HydraGNN/tree/Predictive_GFM_2024

36 MATERIALS SCIENCE↗

Layer-wise Imaging Dataset from Powder Bed Additive Manufacturing Processes for Machine Learning Applications (Peregrine v2021-03)

This dataset contains layer-wise powder bed images from three different powder bed printing technologies – laser powder bed fusion, electron beam powder bed fusion, and binder jetting. This dataset was collected and annotated using the internally-developed Peregrine software tool and is designed primarily to facilitate research into anomaly defect detection using image segmentation or similar techniques. A total of 20 layers are provided for each printing technology, with each layer of data consisting of one or more calibrated images and an annotation file containing pixel-wise ground truth labels. The ground truths were labeled by domain experts, typically printer technicians. Data in this release were collected at Oak Ridge National Laboratory between 2016 and 2020 and were compiled in March 2021.

36 MATERIALS SCIENCE↗

Augmenting Topic Finding in the NASA Aviation Safety Reporting System using Topic Modeling

Context: The NASA Aviation and Safety Reporting System (ASRS) provides various publications to the aviation community (including individual anonymous reports, Callback, Database Search Requests, Directline, and Alerting Messages). Key to these publications are the timely processing of new reports, which is currently done mostly manually by ASRS staff, and which the volume increases yearly. Aim: We investigate whether existing topic modelling techniques are suitable to ease some of the manual effort, and to enhance it with additional visual cues regarding the process of grouping, sense making and labeling incoming (and previous) reports. Method: We evaluate the applicability of WarpLDA topic modelling results combined with three visualization tools, the first two of which have been extended by us in this work for ASRS: Termite, TopicFlow, and LDAVis. Based on the identified limitations in these tools, we propose a methodology for improving them, and evaluate their outputs using ASRS as our test dataset. Results: The user interfaces of Termite, Topicflow and LDAVis were found insufficient for sense-making of the narratives. Moreover, concerns regarding the stability of results due to the inherent randomness of topic modelling, and the lack of a measurable approach for evaluation against the existing ASRS manual workflow were also noted. Conclusion: While many tools to topic modeling and visualization have been proposed, more work is necessary before they can be applied in practical situations to improve existing manual workflows. The methodology presented and applied in this work contribute towards this effort.

ASRS↗

Unsupervised Change Detection for Space Habitats Using 3D Point Clouds

This work presents an algorithm for scene change detection from point clouds to enable autonomous robotic caretaking in future space habitats. Autonomous robotic systems will help maintain future deep-space habitats, such as the Gateway space station, which will be uncrewed for extended periods. Existing scene analysis software used on the International Space Station (ISS) relies on manually-labeled images for detecting changes. In contrast, the algorithm presented in this work uses raw, unlabeled point clouds as inputs. The algorithm first applies modified Expectation-Maximization Gaussian Mixture Model (GMM) clustering to two input point clouds. It then performs change detection by comparing the GMMs using the Earth Mover’s Distance. The algorithm is validated quantitatively and qualitatively using a test dataset collected by an Astrobee robot in the NASA Ames Granite Lab comprising single frame depth images taken directly by Astrobee and full-scene reconstructed maps built with RGB-D and pose data from Astrobee. The runtimes of the approach are also analyzed in depth. The source code is publicly released to promote further development.

robotics↗

Spatial Study 2022: Water Column, Sediment, and Total Ecosystem Respiration Rates across the Yakima River Basin, Washington, USA (v2)

This dataset supports a broader study examining the drivers of spatial variability in sediment respiration rates in the Yakima River Basin and is associated with the manuscript “Sediment-associated processes account for most of the spatial variation in ecosystem respiration in the Yakima River basin” submitted to Nature Communications Earth & Environment (Garayburu-Caruso et al., in review). The dataset provides ecosystem metabolism estimates generated from streamMetabolizer (Appling et al.; 2018) using data collected during the same five-week period at 48 sites within multiple rivers throughout the Yakima River Basin in Washington, USA. Additionally, it includes the scripts used for the analysis and producing the figures in the manuscript. The contents include streamMetabolizer inputs and outputs and additional relevant data needed to generate the main manuscript results. The data included are: total ecosystem respiration, water respiration, calculated sediment-associated respiration, gross primary production outputs from the river corridor model for the Yakima River Basin, median grain size (d50), depth, dissolved oxygen, water temperature, pressure, and annual oxygen consumption. The associated GitHub repository can be found at https://github.com/river-corridors-sfa/SSS_metabolism. Samples collected during this study were labeled as “Second Spatial Study” or “SSS.” Raw time series sensor data, total suspended solids, and depth data from SSS were published at https://data.ess-dive.lbl.gov/datasets/doi:10.15485/1969566. A subset of data from the SSS samples were published in the contiguous United States (CONUS)-Scale Model-Sample (CM) study data package available at https://data.ess-dive.lbl.gov/view/doi:10.15485/1923689 that presents data from across the CONUS. They include dissolved organic carbon (DOC, measured as non-purgeable organic carbon, NPOC), total nitrogen (TN), grain size, aerobic sediment respiration, dissolved oxygen (DO), and temperature. Parent IDs and Site IDs are consistent between the SSS and CM data packages, and they can be mapped directly so data across packages can be used together. Field metadata for the samples in this da This dataset is comprised of one main data folder with four subfolders. The main data folder contains of (1) file-level metadata; (2) data dictionary; (3) total/water column/sediment respiration; (4) gross primary production (GPP); (5) median grain size (d50); and (6) annual oxygen consumption. The “Figures” subfolder contains the figures used in the paper and all intermediate files (including geospatial files). The “Published_Data” contains a readme directing the user to download the public data to reproduce analyses and figures. The “Scripts” folder contains all scripts used in the analyses that were not part of running StreamMetabolizer. Lastly, the “Stream_Metabolizer” folder contains all files associated with running StreamMetabolizer including (1) model input files, (2) model output files, (3) processing scripts, (4) histogram plots of the outputs, and (5) an R project. All files are .csv, .pdf, .R, .Rmd, .Rproj, .html, .png, .txt, .qgz, .cpg, .dbf, .prj, .shp, .shp.ea.iso.xml, .shp.iso.xml, .shx, .sbn. ta package can be found at either link. We acknowledge the Yakama Nation as owners and caretakers of the lands where we collected these data. We thank the Confederated Tribes and Bands of the Yakama Nation Tribal Council and Yakama Nation Fisheries for working with us to facilitate sample collection and optimization of data usage according to their values and worldview.

54 ENVIRONMENTAL SCIENCES↗

Genome-Scale 13 C Fluxomics Modeling for Metabolic Engineering of Saccharomyces cerevisiae .

Synthetic biology is a rapidly developing field that pursues the application of engineering principles and development approaches to biological engineering. Synthetic biology is poised to change the way biology is practiced, and has important practical applications: for example, building genetically engineered organisms to produce biofuels, medicines, and other chemicals. Traditionally, synthetic biology has focused on manipulating a few genes (e.g., in a single pathway or genetic circuit), but its combination with systems biology holds the promise of creating new cellular architectures and constructing complex biological systems from the ground up. Enabling this merge of synthetic and systems biology will require greater predictive capability for modeling the behavior of cellular systems, and more comprehensive data sets for building and calibrating these models. The so-called "-omics" data sets can now be generated via high throughput techniques in the form of genomic, proteomic, transcriptomic, and metabolomic information on the engineered biological system. Of particular interest with respect to the engineering of microbes capable of producing biofuels and other chemicals economically and at scale are metabolomic datasets, and their insights into intracellular metabolic fluxes. Metabolic fluxes provide a rapid and easy to understand picture of how carbon and energy flow throughout the cell. Furthermore, we present a detailed guide to performing metabolic flux analysis and modeling using the open source JBEI Quantitative Metabolic Modeling (jQMM) library. This library allows the user to transform metabolomics data in the form of isotope labeling data from a 13 C labeling experiment into a determination of cellular fluxes that can be used to develop genetic engineering strategies for metabolic engineering.The jQMM library presents a complete toolbox for performing a range of different tasks of interest in metabolic engineering. Various different types of flux analysis and modeling can be performed such as flux balance analysis, 13 C metabolic flux analysis, and two-scale 13 C metabolic flux analysis (2S- 13 C MFA). 2S- 13 C MFA is a novel method that determines genome-scale fluxes without the need of every single carbon transition in the metabolic network. In addition to several other capabilities, the jQMM library can make model based predictions for how various genetic engineering strategies can be incorporated toward bioengineering goals: it can predict the effects of reaction knockouts on metabolism using both the MoMA and ROOM methodologies. In this chapter, we will illustrate the use of the jQMM library through a step-by-step demonstration of flux determination and knockout prediction in a complex eukaryotic model organism: Saccharomyces cerevisiae ( S. cerevisiae ). Included with this chapter is a digital Jupyter Notebook file that provides a computable appendix showing a self-contained example of jQMM usage, which can be changed to fit the user's specific needs. As an open source software project, users can modify and extend the code base to make improvements at will, allowing them to share their development work and contribute back to the jQMM modeling community.

-Omics data↗

Recurrent neural networks for short-term and long-term prediction of geothermal reservoirs

Accurate prediction of geothermal reservoir responses to alternative energy production scenarios is critical for optimizing the development of the underlying resources. While the conventional physics-based models offer a comprehensive prediction tool, data-driven models provide an efficient alternative to build fit-for-purpose predictive models by extracting and using the statistical patterns in the collected data to make predictions. The recurrent neural network (RNN) is a data-driven model that is commonly applied to predict time series sequences. This paper presents a variant of RNN that also utilizes the efficiency of convolutional neural networks (CNN) for the prediction of energy production from geothermal reservoirs. Specifically, a CNN–RNN architecture is developed that takes historical well controls as input (features) and their corresponding production response data as output (labels) to learn an input-output mapping that can predict the future well production responses/performance for any given future well control inputs. The model is paired with a labeling scheme to handle real field disturbances that create data gaps. In addition to the model structure, we introduce a thorough workflow for applying the model, which includes data pre-processing, feature selection, as well as different training strategies for short-term and long-term prediction. Finally, the performance and accuracy of the model are evaluated by applying it to multiple datasets, including a field reservoir model.

15 GEOTHERMAL ENERGY↗

Community Resilience Indicator Analysis: Commonly Used Indicators from Peer-Reviewed Research (Updated for Research Published 2003-2021)

In 2017, FEMA’s National Integration Center (NIC) Technical Assistance (TA) Branch identified a need to establish a data-driven basis for prioritizing locations for TA investment and guiding local emergency management planning. To achieve this goal, FEMA tasked Argonne National Laboratory (Argonne) with identifying commonly used indicators of community resilience across the landscape of published peer-reviewed research. FEMA and Argonne completed the first Community Resilience Indicator Analysis (CRIA) in 2018 and repeated the process in 2022. The CRIA process begins with a literature review and cataloguing of published peer-reviewed assessment methodologies on social vulnerability and community resilience. The literature review findings are then filtered by inclusion criteria established by the CRIA research team to ensure the methodologies are: (1) Quantitative, (2) Data and methodology are publicly available, (3) Calculated at the county level or lower, (4) Examine generalized hazard risk (rather than a singular hazard), and (5) Focused on pre-disaster community conditions. After this, the research team identifies the commonly used indicators across these methodologies and selects the best data source for each indicator. Finally, the research team bins the data for visual display, conducts a correlation analysis and creates a composite index, the FEMA Community Resilience Index (FEMA CRI). In 2018, the CRIA identified eight resilience and vulnerability assessment methodologies and 20 commonly used indicators (indicators used in three or more of the eight methodologies). The FEMA CRI in 2018 was created from these 20 indicators and was produced for at the county level. The 2022 CRIA updated the literature review to expand the list of methodologies examined and followed the same process, resulting in an analysis of 14 methodologies published between 2003 and 2021 and 22 indicators identified as commonly used (indicators used in five or more of the 14 methodologies). In 2022, the research team produced the FEMA CRI at the county and the census tract levels. To make the CRIA data more accessible and more actionable, each individual indicator and the FEMA CRI is binned and included in FEMA’s Resilience Analysis and Planning Tool (RAPT). RAPT enables emergency managers and community partners to quickly visualize relative differences in potential resilience by county, tribe and census tract. By reviewing the data for each of these 22 indicators individually, emergency managers can gain insights for targeted outreach strategies, planning, mitigation investments and response and recovery operations. Communities, regional governments and others can use this data to better understand potential challenges to resilience. As the social science field of examining and validating indicators of resilience evolves, FEMA will update RAPT to provide emergency managers and community partners with additional data and tools to inform planning, mitigation, response and recovery. It is important to understand that the role of the emergency manager is not to change or to “improve” the data, but to plan appropriately for the community characteristics reflected in the data. These datasets are community characteristics that researchers have identified as important considerations for resilience. For example, people with disabilities may have greater challenges to be resilient to disasters. If a community has a high population of people with disabilities, the emergency manager(s) may need to create tailored preparedness outreach programs and strategies to ensure those residents have support if evacuation is necessary. Rather than label these indicators as an absolute measure of resilience, FEMA considers “potential challenges to resilience” a better frame to understand these indicators. Everyone is vulnerable to disasters. While scholars theorize that certain characteristics may make an individual or a household more socially vulnerable, the data does not reflect measures that individuals and/or communities have taken to address potential challenges, such as emergency management planning and outreach or household preparedness measures. To aid emergency managers in understanding how to use these indicators, calling them potential challenges to resilience supports a more positive and strategic application of the data in all phases of emergency management.

99 GENERAL AND MISCELLANEOUS↗

A Data-Driven Framework for Power System Event Type Identification via Safe Semi-Supervised Techniques

Herein this paper investigates the use of phasor measurement unit (PMU) data with deep learning techniques to construct real-time event identification models for transmission networks. Increasing penetration of distributed energy resources represents a great opportunity to achieve decarbonization, as well as challenges in systematic situational awareness. When high-resolution PMU data and sufficient manually recorded event labels are available, the power event identification problem is defined as a statistical classification problem that can be solved by numerous cutting-edge classifiers. However, in real grids, collecting tremendous high-quality event labels is quite expensive. Utilities frequently have a large number of event records without in-depth details (i.e., unlabeled events). To bridge this gap, we propose a novel semi-supervised learning-based method to improve the performance of event classifiers trained with a limited number of labeled events by exploiting the information from massive unlabeled events. In other words, compared to existing data-driven methods, our method requires only a small portion of labeled data to achieve a similar level of accuracy. Meanwhile, this work discusses and addresses the performance degradation caused by class distribution mismatch between the training set and the real applications. Based on the proposed safe learning mechanism, our model does not directly use all unlabeled events during model training, but selectively uses them through a comprehensive evaluation procedure. Numerical studies on a sizable PMU dataset have been used to validate the performance of the proposed method.

42 ENGINEERING↗

Effect of image resolution on automated classification of chest X-rays

Deep learning (DL) models have received much attention lately for their ability to achieve expert-level performance on the accurate automated analysis of chest X-rays (CXRs). Recently available public CXR datasets include high resolution images, but state-of-the-art models are trained on reduced size images due to limitations on graphics processing unit memory and training time. As computing hardware continues to advance, it has become feasible to train deep convolutional neural networks on high-resolution images without sacrificing detail by downscaling. This study examines the effect of increased resolution on CXR classification performance. We used the publicly available MIMIC-CXR-JPG dataset, comprising 377,110 high resolution CXR images for this study. We applied image downscaling from native resolution to 2048 × 2048 pixels, 1024 × 1024 pixels, 512 × 512 pixels, and 256 × 256 pixels and then we used the DenseNet121 and EfficientNet-B4 DL models to evaluate clinical task performance using these four downscaled image resolutions. We find that while some clinical findings are more reliably labeled using high resolutions, many other findings are actually labeled better using downscaled inputs. We qualitatively verify that tasks requiring a large receptive field are better suited to downscaled low resolution input images, by inspecting effective receptive fields and class activation maps of trained models. Lastly, we show that stacking an ensemble across resolutions outperforms each individual learner at all input resolutions while providing interpretable scale weights, indicating that diverse information is extracted across resolutions.

47 OTHER INSTRUMENTATION↗

Big Data Synchrophasor Monitoring and Analytics for Resiliency Tracking (BDSMART)

This report contains key findings from a project titled Big Data Synchrophasor Monitoring and Analytics for Resiliency Tracking (BDSMART), which was carried out through a collaborative effort of a team of researchers from Texas A&M Engineering Experiment Station, Temple University, and Quanta Technology, LLC. The in-kind support came from OSIsoft (acquired by AVEVA), which provided their PI Historian software to demonstrate the use case of streaming PMU data. The first section of the report describes the project goals and objectives related to the development of Machine Learning (ML) models capable of detecting and classifying events by processing phasor measurements captured in the field by Phasor Measurement Units (PMUs). The data for this study was contributed by the utilities/ISOs from the Western and Eastern interconnects and ERCOT, further referred to as Interconnect B (IC B), Interconnect A (IC A), and Interconnect C (IC C), respectively. The approach that the BDSMART Research Team proposed and the key research tasks defined by the team are outlined in this section. The next section describes the technical approach. We first discuss the data constraints related to the PMU measurements and data interpretation constraints imposed by the data contributors. They provided neither the topological information of the grid nor PMU placement locations and captured recorded data at very few locations in the system with the reporting rate of either 30 or 60 fps. The recordings are mostly positive sequence voltage, frequency, and ROCOF, and in some limited cases, three-phase voltages and currents. We then reflect on the bad data issues that stem from poor recording practices and vague definitions of the PMU status bits to supposedly be used for bad data identification. Finally, the data discovery points to imprecise time stamps with incomplete event start/end time, as well as inconsistent and incomplete event labeling, which combined make the implementation of the data models using supervising learning quite challenging. Following the data discovery study, we hypothesize that because the IC B data has the most complete labels, we should focus our model development on that data and then test it on data from other interconnects. We also define the common metrics used to evaluate the results from the ML algorithm tests. We concluded this section by summarizing the common ML models we used and explaining how we implemented and tested them. The issues from this section are expanded in the Training Dataset Report from this project. The final section of this report deals with the accomplishments and conclusions. As the accomplishments, we formulate the problem we are solving and what is achieved by solving the problem. We then reflect on each of the analytics tools we developed and point out the performance of each tool when applied to solving the mentioned problems. We reference this work for further details to the papers we published on each tool. In the conclusions, we give recommendations on how to improve future PMU recording practices to facilitate the ML algorithm implementation and guidance for the future standardization work aimed at clarifying the ambiguities associated with the PMU status bits. We finally list future tasks that can bring about further improvements in the proposed algorithms. The issues from this section are expanded in the Training, and Test Dataset Report filed at the project completion date.

97 MATHEMATICS AND COMPUTING↗

Comparative differential cuproproteomes of Rhodobacter capsulatus reveal novel copper homeostasis related proteins

Abstract Copper (Cu) is an essential, but toxic, micronutrient for living organisms and cells have developed sophisticated response mechanisms towards both the lack and the excess of Cu in their environments. In this study, we achieved a global view of Cu-responsive changes in the prokaryotic model organism Rhodobacter capsulatus using label-free quantitative differential proteomics. Semi-aerobically grown cells under heterotrophic conditions in minimal medium (∼0.3 μM Cu) were compared with cells supplemented with either 5 μM Cu or with 5 mM of the Cu-chelator bathocuproine sulfonate. Mass spectrometry based bottom-up proteomics of unfractionated cell lysates identified 2430 of the 3632 putative proteins encoded by the genome, producing a robust proteome dataset for R. capsulatus. Use of biological and technical replicates for each growth condition yielded high reproducibility and reliable quantification for 1926 of the identified proteins. Comparison of cells grown under Cu-excess or Cu-depleted conditions to those grown under minimal Cu-sufficient conditions revealed that 75 proteins exhibited statistically significant (p < 0.05) abundance changes, ranging from 2- to 300-fold. A subset of the highly Cu-responsive proteins was orthogonally probed using molecular genetics, validating that several of them were indeed involved in cellular Cu homeostasis.

59 BASIC BIOLOGICAL SCIENCES↗

Harnessing large language models’ zero-shot and few-shot learning capabilities for regulatory research

Abstract Large language models (LLMs) are sophisticated AI-driven models trained on vast sources of natural language data. They are adept at generating responses that closely mimic human conversational patterns. One of the most notable examples is OpenAI's ChatGPT, which has been extensively used across diverse sectors. Despite their flexibility, a significant challenge arises as most users must transmit their data to the servers of companies operating these models. Utilizing ChatGPT or similar models online may inadvertently expose sensitive information to the risk of data breaches. Therefore, implementing LLMs that are open source and smaller in scale within a secure local network becomes a crucial step for organizations where ensuring data privacy and protection has the highest priority, such as regulatory agencies. As a feasibility evaluation, we implemented a series of open-source LLMs within a regulatory agency’s local network and assessed their performance on specific tasks involving extracting relevant clinical pharmacology information from regulatory drug labels. Our research shows that some models work well in the context of few- or zero-shot learning, achieving performance comparable, or even better than, neural network models that needed thousands of training samples. One of the models was selected to address a real-world issue of finding intrinsic factors that affect drugs' clinical exposure without any training or fine-tuning. In a dataset of over 700 000 sentences, the model showed a 78.5% accuracy rate. Our work pointed to the possibility of implementing open-source LLMs within a secure local network and using these models to perform various natural language processing tasks when large numbers of training examples are unavailable.

Biochemistry & Molecular Biology↗

Soil and Water Chemistry and Trace Metal Extractability and Speciation in Wetland Soils from Illinois and South Carolina and Stream Sediments from Tennessee

Dataset revised on October 15, 2021. This revision adds sulfur and iron X-ray absorption near-edge structure spectra for the wetland soils and stream sediments from the field areas. It also renames the sample locations in a way that is more intuitive to readers of the companion paper that is under review. Finally, the data filenames and organization have been updated in their labeling to parallel the data sources in the associated paper. The abstract text and methods were also revised to reflect the data that was added to the dataset.Trace metals are essential for microbially-mediated biogeochemical processes occurring in anoxic wetland soils and stream bed sediments, such as denitrification, methanogenesis, and mercury methylation. Low availability of these elements may potentially inhibit key components of anaerobic carbon and nitrogen cycling and contaminant transformation. The solid-phase speciation of trace metals likely plays an important role in controlling their bioavailability. Metal speciation is well studied in contaminated soils and sediments as well as those naturally elevated in trace metals. However, less is known regarding the chemical forms of trace metals in systems having concentrations similar to geological background levels, the very settings where metal limitations may be most prevalent. We have investigated trace metal concentrations, extractability, and solid-phase speciation in three freshwater subsurface aquatic systems: marsh wetland soils, riparian wetland soils, and the sediments of a streambed.Data are provided for marsh wetland soils at Argonne National Laboratory, riparian wetland soils in the Tims Branch watershed at Savannah River National Laboratory, and stream bed sediments from East Fork Poplar Creek near Oak Ridge National Laboratory. Soil and sediment elemental abundances, mineralogy, and extractable nutrients as well as dissolved major elements, anions, trace metals, and nutrients in the overlying surface waters are provided. In addition, the results of sequential chemical extraction for the trace metals cobalt, nickel, copper, and zinc from the soils and sediment are reported as well as X-ray absorption near-edge structure (XANES) spectra in these materials are reported. To aid interpretation of these data, XANES spectra of sulfur in the soils and sediments as well as both XANES and extended X-ray absorption fine structure (EXAFS) spectra of iron in these materials are reported. The data package also includes the XANES spectra of reference standards and a potential interferent in the measurements. All data are provided in text-based CSV format with header sections indicating the data contained in each file and the corresponding units. Note that "u" is used in place of Greek lower case mu to indicate the micro prefix on units.

54 ENVIRONMENTAL SCIENCES↗