Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “small dataset”

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 253 records · Page 14

High-Mileage Courier Fleet Vehicle Laboratory Battery Pack Testing

For one of each of the AVTA's plug-in hybrid electric vehicles, battery electric vehicles, and some hybrid electric vehicles tested in high-mileage courier fleets, the high-voltage traction battery packs were removed from the vehicle and tested at the beginning and end of fleet testing. For some vehicles, batteries also were tested at periodic intervals during fleet testing. Standard reference performance tests were conducted to characterize battery degradation over time. This dataset contains results from two or more rounds of battery tests for 21 distinct year/make/model vehicles (see reference ["INL Advanced Vehicle Testing Activity: On-road Logger and Laboratory Battery Pack Testing Vehicle List"](https://avt.inl.gov/sites/default/files/pdf/reports/DatasetVehicleList.pdf) for full list of vehicles). Each round of battery testing included the "Static Capacity Test" and the "Hybrid Pulse Power Characterization (HPPC) Test", conducted according to test procedures published in the United States Advanced Battery Consortium ["Battery Test Manual For Power-Assist Hybrid Electric Vehicles"](https://www.uscar.org/commands/files_download.php?files_id=57), ["Battery Test Manual For Plug-In Hybrid Electric Vehicles"](https://www.uscar.org/commands/files_download.php?files_id=168), and ["Electric Vehicle Battery Test Procedures Manual"](https://www.uscar.org/commands/files_download.php?files_id=5) prior to the time of testing. These tests were performed by Intertek Testing Services, North America. (This dataset will be shared by API; a small sample of the vehicle battery and test data has been extracted and is available for download while the API is being developed.)

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

SAIL-Net Raw and Post Corrected POPS Data Fall 2021 - Summer 2023

SAIL-Net is a DOE funded project in the East River Watershed near Crested Butte, Colorado with the goal of advancing our understanding of aerosol-cloud interactions in complex, mountainous regions. Through the deployment of a network of six low cost microphysics nodes in Fall 2021 in the same domain at the SAIL campaign, SAIL-Net provides data on aerosol size distributions, cloud condensation nuclei (CCN), and ice nucleation particles (INP). This network enables the investigation of small-scale variations in complex terrain. Two datasets are provided - one containing raw data and the other containing post-corrected data. The raw dataset provides the raw data recorded from the POPS which were deployed at each of the six sites. These data are organized by site and broken down into daily data files. The six site names used here are: “gothic”, “irwin”, “cbtop”, “cbmid”, “pumphouse”, and “snodgrass”. These data are not cleaned or post-corrected, but some flags have been added. The data are reported at 1 second time resolution. The post-corrected dataset provides the post-corrected and cleaned data recorded from the POPS which were deployed at each of the six sites. This data are also organized by site (same as those found in the raw data) and broken down into daily data files. Unlike the raw POPS data, these data have already been cleaned to remove what we believe are bad values. These data should be ready to use with no cleaning. For a full description of the cleaning and post-correction process, see the readme.

54 ENVIRONMENTAL SCIENCES↗

Windowed and Wavelet Analysis of Marine Stratocumulus Cloud Inhomogeneity

To improve radiative transfer calculations for inhomogeneous clouds, a consistent means of modeling inhomogeneity is needed. One current method of modeling cloud inhomogeneity is through the use of fractal parameters. This method is based on the supposition that cloud inhomogeneity over a large range of scales is related. An analysis technique named wavelet analysis provides a means of studying the multiscale nature of cloud inhomogeneity. In this paper, the authors discuss the analysis and modeling of cloud inhomogeneity through the use of wavelet analysis. Wavelet analysis as well as other windowed analysis techniques are used to study liquid water path (LWP) measurements obtained during the marine stratocumulus phase of the First ISCCP (International Satellite Cloud Climatology Project) Regional Experiment. Statistics obtained using analysis windows, which are translated to span the LWP dataset, are used to study the local (small scale) properties of the cloud field as well as their time dependence. The LWP data are transformed onto an orthogonal wavelet basis that represents the data as a number of times series. Each of these time series lies within a frequency band and has a mean frequency that is half the frequency of the previous band. Wavelet analysis combined with translated analysis windows reveals that the local standard deviation of each frequency band is correlated with the local standard deviation of the other frequency bands. The ratio between the standard deviation of adjacent frequency bands is 0.9 and remains constant with respect to time. This ratio defined as the variance coupling parameter is applicable to all of the frequency bands studied and appears to be related to the slope of the data's power spectrum. Similar analyses are performed on two cloud inhomogeneity models, which use fractal-based concepts to introduce inhomogeneity into a uniform cloud field. The bounded cascade model does this by iteratively redistributing LWP at each scale using the value of the local mean. This model is reformulated into a wavelet multiresolution framework, thereby presenting a number of variants of the bounded cascade model. One variant introduced in this paper is the 'variance coupled model,' which redistributes LWP using the local standard deviation and the variance coupling parameter. While the bounded cascade model provides an elegant two- parameter model for generating cloud inhomogeneity, the multiresolution framework provides more flexibility at the expense of model complexity. Comparisons are made with the results from the LWP data analysis to demonstrate both the strengths and weaknesses of these models.

Gollmer, Steven M.↗

A Lightweight I/O Scheme to Facilitate Spatial and Temporal Queries of Scientific Data Analytics

In the era of petascale computing, more scientific applications are being deployed on leadership scale computing platforms to enhance the scientific productivity. Many I/O techniques have been designed to address the growing I/O bottleneck on large-scale systems by handling massive scientific data in a holistic manner. While such techniques have been leveraged in a wide range of applications, they have not been shown as adequate for many mission critical applications, particularly in data post-processing stage. One of the examples is that some scientific applications generate datasets composed of a vast amount of small data elements that are organized along many spatial and temporal dimensions but require sophisticated data analytics on one or more dimensions. Including such dimensional knowledge into data organization can be beneficial to the efficiency of data post-processing, which is often missing from exiting I/O techniques. In this study, we propose a novel I/O scheme named STAR (Spatial and Temporal AggRegation) to enable high performance data queries for scientific analytics. STAR is able to dive into the massive data, identify the spatial and temporal relationships among data variables, and accordingly organize them into an optimized multi-dimensional data structure before storing to the storage. This technique not only facilitates the common access patterns of data analytics, but also further reduces the application turnaround time. In particular, STAR is able to enable efficient data queries along the time dimension, a practice common in scientific analytics but not yet supported by existing I/O techniques. In our case study with a critical climate modeling application GEOS-5, the experimental results on Jaguar supercomputer demonstrate an improvement up to 73 times for the read performance compared to the original I/O method.

Temporal Queries↗

Broad Line Radio Galaxies Observed with Fermi-LAT: The Origin of the GeV Gamma-Ray Emission

We report on a detailed investigation of the {gamma}-ray emission from 18 broad line radio galaxies (BLRGs) based on two years of Fermi Large Area Telescope (LAT) data. We confirm the previously reported detections of 3C 120 and 3C 111 in the GeV photon energy range; a detailed look at the temporal characteristics of the observed {gamma}-ray emission reveals in addition possible flux variability in both sources. No statistically significant {gamma}-ray detection of the other BLRGs was however found in the considered dataset. Though the sample size studied is small, what appears to differentiate 3C 111 and 3C 120 from the BLRGs not yet detected in {gamma}-rays is the particularly strong nuclear radio flux. This finding, together with the indications of the {gamma}-ray flux variability and a number of other arguments presented, indicate that the GeV emission of BLRGs is most likely dominated by the beamed radiation of relativistic jets observed at intermediate viewing angles. In this paper we also analyzed a comparison sample of high accretion-rate Seyfert 1 galaxies, which can be considered radio-quiet counterparts of BLRGs, and found none were detected in {gamma}-rays. A simple phenomenological hybrid model applied for the broad-band emission of the discussed radio-loud and radio-quiet type 1 active galaxies suggests that the relative contribution of the nuclear jets to the accreting matter is {ge} 1% on average for BLRGs, while {le} 0.1% for Seyfert 1 galaxies.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Generalized representative structures for atomistic systems

A new method is presented to generate atomic structures that reproduce the essential characteristics of arbitrary material systems, phases, or ensembles. Previous methods allow one to reproduce the essential characteristics (e.g. the chemical disorder) of a large random alloy within a small crystal structure. The ability to generate small representations of random alloys, along with the restriction to crystal systems, results from using the fixed-lattice cluster correlations to describe structural characteristics. A more general description of the structural characteristics of atomic systems is obtained using complete sets of atomic environment descriptors. These are used within for generating representative atomic structures without restriction to fixed lattices. A general data-driven approach is provided here utilizing the atomic cluster expansion (ACE) basis. The N-body ACE descriptors are a complete set of atomic environment descriptors that span both chemical and spatial degrees of freedom and are used within for describing atomic structures. The generalized representative structure (GRS) method presented within generates small atomic structures that reproduce ACE descriptor distributions corresponding to arbitrary structural and chemical complexity. It is shown that systematically improvable representations of crystalline systems on fixed parent lattices, amorphous materials, liquids, and ensembles of atomic structures may be produced efficiently through optimization algorithms. With the GRS method, we highlight reduced representations of atomistic machine-learning training datasets that contain similar amounts of information and small 40–72 atom representations of liquid phases. The ability to use GRS methodology as a driver for informed novel structure generation is also demonstrated. The advantages over other data-driven methods and state-of-the-art methods restricted to high-symmetry systems are highlighted.

atomic cluster expansion↗

HydraGNN_Predictive_GFM_2026 - Ensemble of predictive graph foundation models for atomistic materials modeling

This release contains data and parameters of HydraGNN-based graph foundation models trained as a result of the work published in the pre-print "Exascale Multi-Task Graph Foundation Models for Imbalanced, Multi-Fidelity Atomistic Data" by M. Lupo Pasini et al. (https://arxiv.org/abs/2604.15380). We jointly train on 16 open first-principles datasets (544+ million structures covering 85+ elements) using a multi-task architecture with per-dataset heads and a scalable ADIOS2/DDStore data pipeline. On Frontier, we execute six large-scale DeepHyper hyperparameter optimization campaigns in FP64 and promote the top-performing message-passing models to sustained 2,048-node training, yielding a PaiNN-based lead model. The version of HydraGNN used to generate the outputs provided in this release is HydraGNN v5.0 (https://github.com/ORNL/HydraGNN/releases/tag/v5.0) The list of datasets used for the training of the graph foundation model is the following: 1) Alexandria [1] 2) ANI1x [2] 3) MPTrj [3] 4) Open Catalyst 2020 (OC20) [4] 5) Open Catalyst 2022 (OC22) [5] 6) Open Catalyst 2025 (OC25) [6] 7) Open Direct ir Capture 2023 (ODAC23) [7] 8) Open Materials 2024 (OMat24) [8] 9) Open Molecules 2025 (OMol25) [9] 10) OMol25-neutral (subset of OMol25 that contains only molecules with zero total charge) 11) OMol25-non-neutral (subset of OMol25 that contains only molecules with non-zero total charge) 12) Open Polymers 2026 (OPoly2026) [10] 13) Nabla2DFT [11] 14) QCML [12] 15) QM7X [reference 13] 16) transition1x [14] Dataset references: [1] J. Schmidt et al., “A dataset of 175k stable and metastable materials calculated with the PBEsol and SCAN functionals,” Scientific Data, vol. 9, p. 64, 2022. [2] J. S. Smith et al., “The ANI-1ccx and ANI-1x data sets, coupled-cluster and density functional theory properties for molecules,” Scientific Data, vol. 7, p. 134, 2020. [Online]. Available: https: //www.nature.com/articles/s41597-020-0473-z [3] A. Jain et al., “Commentary: The Materials Project: A materials genome approach to accelerating materials innovation,” APL Materials, vol. 1, no. 1, p. 011002, 07 2013. [Online]. Available: https://doi.org/10.1063/1.4812323 [4] L. Chanussot et al., “Open catalyst 2020 (oc20) dataset and community challenges,” ACS Catalysis, vol. 11, no. 10, pp. 6059–6072, 2021. [Online]. Available: https://doi.org/10.1021/acscatal.0c04525 [5] K. Tran et al., “Open catalyst 2022 (oc22) dataset and challenges for oxidation electrocatalysts,” ACS Catalysis, vol. 13, no. 5, pp. 3066–3084, 2023. [Online]. Available: https://doi.org/10.1021/acscatal.2c05426 [6] S. J. Sahoo et al., “The open catalyst 2025 (oc25) dataset and models for solid-liquid interfaces,” arXiv preprint arXiv:2509.17862, 2025. [Online]. Available: https://arxiv.org/abs/2509.17862 [7] A. Sriram et al., “The open DAC 2023 dataset and challenges for sorbent discovery in direct air capture,” ACS Central Science, vol. 10, no. 5, pp. 923–941, 2024. [8] L. Barroso-Luque et al., “Open materials 2024 (omat24) inorganic materials dataset and models,” 2024. [Online]. Available: https://arxiv.org/abs/2410.12771 [9] D. S. Levine et al., “The open molecules 2025 (OMol25) dataset, evaluations, and models,” 2025. [Online]. Available: https://arxiv.org/abs/2505.08762 [10] D. S. Levine et al., The open polymers 2026 (OPoly26) dataset and evaluations,” arXiv preprint arXiv:2512.23117, 2025. [Online]. Available: https://arxiv.org/abs/2512.23117 [11] K. Khrabrov et al., “Nabla2dft: A universal quantum chemistry dataset of drug-like molecules and a benchmark for neural network potentials,” in NeurIPS 2024 Datasets and Benchmarks Track, 2024. [Online]. Available: https://openreview.net/forum?id=ElUrNM9U8c [12] S. Ganscha et al., “The QCML dataset, quantum chemistry reference data from 33.5M DFT and 14.7B semi-empirical calculations,” Scientific Data, vol. 12, p. 406, 2025. [13] J. Hoja et al., “QM7-X, a comprehensive dataset of quantum-mechanical properties spanning the chemical space of small organic molecules,” Scientific Data, vol. 8, p. 43, 2021. [Online]. Available: https://www.nature.com/articles/s41597-021-00812-2 [14] M. Schreiner et al., “Transition1x - a dataset for building generalizable reactive machine learning potentials,” Scientific Data, vol. 9, p. 779, 2022. The folder "datasets_ADIOS2_format" 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 "datasets_ADIOS2_format" directory contains 2 sub-directories, one for the version "v1" of the datasets and one for the version "v2" of the datasets. The version "v1" of the datasets provides values of the total energy as they are extracted from the original data as it was released by the respective institutions. The version "v2" of the datasets provides values of the energy that have been realigned. The realignment was performed by training a linear regression model that predicts the total energy as a function of the chemical composition of the atomistic structure, and then subtract such prediction from the original value of the total energy. Both folders "v1" and "v2" contain 16 sub-directories, each corresponding to an ADIOS2-formatted dataset The folder "DeepHyper-results" contains the configurational files and model's parameters for all the 186 HPO trials that were successfully completed by the scalable hyperparameter optimization (HPO) runs on Frontier. The content of the folder "DeepHyper-results" I structured as follows: 1) task-list.txt: list of mpnn name, jobid, and deephyper task id 2) gfm_${MPNN}_${JOBID}_0.${TASKID}: run directory with checkpoint files 3) gfm_${MPNN}: deephyper summary directory (*.csv) for each specific MPNN type 4) deephyper-experiment-${JOBID}: output and error logs for each job The file "deephyper-sorted.csv" contains the details of each HydraGNN model built and tested by HPO, obtained by merging the (*.csv) filed from each HPO run executed. Out of all the HPO trials, we selected 10 to continue the training of the respective HydraGNN models. Due to limited computational budget available in the LRN070 allocation we could not complete the training till convergence for all these 10 selected models. The folder "models" contains multiple sub-folders, one per each HydraGNN model trained. Each model sub-folder contains the parameters of each HydraGNN model, with multiple checkpoint-restarts. The list of sub-folders are as follows: 1) multidataset_hpo-BEST1-fp64 2) multidataset_hpo-BEST2-fp64 3) multidataset_hpo-BEST3-fp64 4) multidataset_hpo-BEST4-fp64 5) multidataset_hpo-BEST5-fp64 6) multidataset_hpo-BEST6-fp64 7) multidataset_hpo-BEST7-fp64 8) multidataset_hpo-BEST8-fp64 9) multidataset_hpo-BEST9-fp64 10) multidataset_hpo-BEST10-fp64 Within each one of these folders, additional auxiliary log files are provided with descriptions about how the training proceeded. The lead PaiNN-model is contained inside "multidataset_hpo-BEST6-fp64". The file "mlp_branch_weights" contains the parameters of the multi-layer perceptron (MLP) used to reconcile the predictions of the 16 output decoding heads of the HydragNN architectures. The MLP takes in input the chemical composition of the atomistic structure and predicts averaging weights to linearly mix the predictions of each output decoding head toward consolidating them into a single one. The folder "1.1billion-structure-inference" contains 1.1 billion atomistic structures randomly generated. Each structures is associated with energy and forces predicted with the lead-PaiNN model combined with the MLP model for reconciliation of the multi-branch predictions generated by the 16 output decoding heads. The folder "1.1billion-structure-inference" contains 9,300 (*.tar.gz) subdirectories, one per Frontier compute node used to execute the inference at exascale. Once uncompressed, each (*.tar.gz) subdirectory contains an ADIOS2 (*.bp) file container, where each atomistic structure is stored as a PyTorch-Geometric Data object. The file "export_dataset_environment_variables.sh" contains the environment variables that need to be set before running the HydraGNN code to reproduce the results provided in this dataset release. The code that can be used to load the ADIOS2 files, load HydraGNN models, and run inference is available at: https://github.com/ORNL/HydraGNN/releases/tag/v5.0

36 MATERIALS SCIENCE↗

Airfoil Computational Fluid Dynamics - 2k shapes, 25 AoA's, 3 Re numbers

This dataset contains aerodynamic quantities - including flow field values (momentum, energy, and vorticity) and summary values (coefficients of lift, drag, and momentum) - for 1,830 airfoil shapes computed using the HAM2D CFD (computational fluid dynamics) model. The airfoil shapes were designed using the separable shape tensor parameterization that encodes two-dimensional shapes as elements of the Grassmann manifold. This data-driven approach learns two independent spaces of parameter from a collection of sample airfoils. The first captures large-scale, linear perturbations, and the second defines small-scale, higher-order perturbations. For this dataset, we used the G2Aero database of over 19,000 airfoil shapes to learn a parameter space that captured a wide array of shape characteristics. We sampled airfoil designs over both parameter spaces to explore the full range of possible shape variations. The aerodynamic quantities for the generated airfoil were obtained using the HAM2D code, which is a finite-volume Reynolds-averaged Navier-Stokes (RANS) flow solver. We employ a fifth-order WENO scheme for spatial reconstruction with Roe's flux difference scheme for inviscid flux and second-order central differencing for viscous flux. A preconditioned GMRES method is applied for implicit integration. The Spalart-Allmaras 1-eq turbulence model is used for the turbulence closure, and the Medida-Baeder 2-eq transition model is applied to account for the effects of laminar turbulent transition. The airfoil grid is generated with a total of 400 points on the airfoil surface, the initial wall-normal spacing of y+ = 1, and an outer boundary located at 300 chord lengths away from the wall. The CFD simulations are performed at a freestream Mach number of 0.1, for or three different Reynolds' numbers (3M, 6M, and 9M), and for 25 angles of attack from -4 deg. to 20 deg. with 1 degree increments. Across all these various parameters, this dataset includes the results from over 250,000 CFD simulations. The simulations were performed using the Bridges-2 system at the Pittsburgh Supercomputing Center in February 2023 as part of the INTEGRATE project funded by the Advanced Research Projects Agency - Energy, in the U.S. Department of Energy. The data was collected, reformatted, and preprocessed for this OEDI submission in July 2023 under the Foundational AI for Wind Energy project funded by the U.S. Department of Energy Wind Energy Technologies Office. This dataset is intended to serve as a benchmark against which new artificial intelligence (AI) or machine learning (ML) tools may be tested. Baseline AI/ML methods for analyzing this dataset have been implemented, and a link to their repository containing those models has been provided. The .h5 data file structure can be found in the GitHub Repository resource under explore_airfoil_2k_data.ipynb.

2k↗

Artificial neural network approach for multiphase segmentation of battery electrode nano-CT images

The segmentation of tomographic images of the battery electrode is a crucial processing step, which will have an additional impact on the results of material characterization and electrochemical simulation. However, manually labeling X-ray CT images (XCT) is time-consuming, and these XCT images are generally difficult to segment with histographical methods. We propose a deep learning approach with an asymmetrical depth encode-decoder convolutional neural network (CNN) for real-world battery material datasets. This network achieves high accuracy while requiring small amounts of labeled data and predicts a volume of billions voxel within few minutes. While applying supervised machine learning for segmenting real-world data, the ground truth is often absent. The results of segmentation are usually qualitatively justified by visual judgement. We try to unravel this fuzzy definition of segmentation quality by identifying the uncertainty due to the human bias diluted in the training data. Further CNN trainings using synthetic data show quantitative impact of such uncertainty on the determination of material’s properties. Nano-XCT datasets of various battery materials have been successfully segmented by training this neural network from scratch. We will also show that applying the transfer learning, which consists of reusing a well-trained network, can improve the accuracy of a similar dataset.

25 ENERGY STORAGE↗

Collaborative project. Ocean-atmosphere interaction from meso- to planetary-scale. Mechanics, parameterization, and variability

Most climate models are currently run with grid spacings of around 100km, which, with today’s computing power, allows for long (up to 1000 year) simulations, or ensembles of simulations to explore climate change and variability. However this grid spacing does not resolve important components of the weather/climate system such as atmospheric fronts and mesoscale systems, and ocean boundary currents and eddies. The overall aim of this project has been to look at the effect of these small-scale features on the weather/climate system using a suite of high and low resolution climate models, idealized models and observations. High-resolution global coupled integrations using CAM/CESM were carried out at NCAR by the lead PI. At TAMU, we have complemented the work at NCAR by analyzing datasets from the high-resolution (28km) CESM integrations (Small et al., 2014) as well as very high resolution (9km, 3km) runs using a coupled regional climate (CRCM) carried out locally. The main tasks carried out were: 1. Analysis of surface wind in observations and high-resolution CAM/CCSM simulations 2. Development of a feature-tracking algorithm for studying midlatitude air-sea interaction by following oceanic mesoscale eddies and creating composites of the atmospheric response overlying the eddies. 3. Applying the Lagrangian analysis technique in the Gulf Stream region to compare data from observational reanalyses, global CESM coupled simulations, 9km regional coupled simulations and 3km convection-resolving regional coupled simulations. Our main findings are that oceanic mesoscale eddies influence not just the atmospheric boundary layer above them, but also the lower portions of the free troposphere above the boundary layer. Such a vertical response could have implications for a remote influence of Gulf Stream oceanic eddies on North Atlantic weather patterns through modulation of the storm track, similar to what has been noted in the North Pacific. The coarse resolution observational reanalyses perhaps underestimate the atmospheric response, but the 28km global model resolution appears to be adequate to capture some, but not all, aspects of the boundary response. The higher resolution regional models show a stronger response in certain fields such as the latent heat flux.

58 GEOSCIENCES↗

Re-Organizing Earth Observation Data Storage to Support Temporal Analysis of Big Data

The Earth Observing System Data and Information System archives many datasets that are critical to understanding long-term variations in Earth science properties. Thus, some of these are large, multi-decadal datasets. Yet the challenge in long time series analysis comes less from the sheer volume than the data organization, which is typically one (or a small number of) time steps per file. The overhead of opening and inventorying complex, API-driven data formats such as Hierarchical Data Format introduces a small latency at each time step, which nonetheless adds up for datasets with O(10^6) single-timestep files. Several approaches to reorganizing the data can mitigate this overhead by an order of magnitude: pre-aggregating data along the time axis (time-chunking); storing the data in a highly distributed file system; or storing data in distributed columnar databases. Storing a second copy of the data incurs extra costs, so some selection criteria must be employed, which would be driven by expected or actual usage by the end user community, balanced against the extra cost.

data storage↗

Too big, too small, or just right? A benchmark assessment of density functional theory for predicting the spatial extent of the electron density of small chemical systems

Multipole moments are the first-order responses of the energy to spatial derivatives of the electric field strength. The quality of density functional theory prediction of molecular multipole moments thus characterizes errors in modeling the electron density itself, as well as the performance in describing molecules interacting with external electric fields. However, only the lowest non-zero moment is translationally invariant, making the higher-order moments origin-dependent. Therefore, instead of using the 3 × 3 quadrupole moment matrix, we utilize the translationally invariant 3 × 3 matrix of second cumulants (or spatial variances) of the electron density as the quantity of interest (denoted by K). The principal components of K are the square of the spatial extent of the electron density along each axis. A benchmark dataset of the principal components of K for 100 small molecules at the coupled cluster singles and doubles with perturbative triples at the complete basis set limit is developed, resulting in 213 independent K components. The performance of 47 popular and recent density functionals is assessed against this Var213 dataset. Several functionals, especially double hybrids, and also SCAN and SCAN0 predict reliable second cumulants, although some modern, empirically parameterized functionals yield more disappointing performance. The H, Li, and Be atoms, in particular, are challenging for nearly all methods, indicating that future functional development could benefit from the inclusion of their density information in training or testing protocols.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

SU-E-T-630: Predictive Modeling of Mortality, Tumor Control, and Normal Tissue Complications After Stereotactic Body Radiotherapy for Stage I Non-Small Cell Lung Cancer

Purpose: While rates of local control have been well characterized after stereotactic body radiotherapy (SBRT) for stage I non-small cell lung cancer (NSCLC), less data are available characterizing survival and normal tissue toxicities, and no validated models exist assessing these parameters after SBRT. We evaluate the reliability of various machine learning techniques when applied to radiation oncology datasets to create predictive models of mortality, tumor control, and normal tissue complications. Methods: A dataset of 204 consecutive patients with stage I non-small cell lung cancer (NSCLC) treated with stereotactic body radiotherapy (SBRT) at the University of Pennsylvania between 2009 and 2013 was used to create predictive models of tumor control, normal tissue complications, and mortality in this IRB-approved study. Nearly 200 data fields of detailed patient- and tumor-specific information, radiotherapy dosimetric measurements, and clinical outcomes data were collected. Predictive models were created for local tumor control, 1- and 3-year overall survival, and nodal failure using 60% of the data (leaving the remainder as a test set). After applying feature selection and dimensionality reduction, nonlinear support vector classification was applied to the resulting features. Models were evaluated for accuracy and area under ROC curve on the 81-patient test set. Results: Models for common events in the dataset (such as mortality at one year) had the highest predictive power (AUC = .67, p < 0.05). For rare occurrences such as radiation pneumonitis and local failure (each occurring in less than 10% of patients), too few events were present to create reliable models. Conclusion: Although this study demonstrates the validity of predictive analytics using information extracted from patient medical records and can most reliably predict for survival after SBRT, larger sample sizes are needed to develop predictive models for normal tissue toxicities and more advanced machine learning methodologies need be consider in the future.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

A Nomogram to Predict Radiation Pneumonitis, Derived From a Combined Analysis of RTOG 9311 and Institutional Data

Purpose: To test the Washington University (WU) patient dataset, analysis of which suggested that superior-to-inferior tumor position, maximum dose, and D35 (minimum dose to the hottest 35% of the lung volume) were valuable to predict radiation pneumonitis (RP), against the patient database from Radiation Therapy Oncology Group (RTOG) trial 9311. Methods and Materials: The entire dataset consisted of 324 patients receiving definitive conformal radiotherapy for non-small-cell lung cancer (WU = 219, RTOG 9311 = 129). Clinical, dosimetric, and tumor location parameters were modeled to predict RP in the individual datasets and in a combined dataset. Association quality with RP was assessed using Spearman's rank correlation (r) for univariate analysis and multivariate analysis; comparison between subgroups was tested using the Wilcoxon rank sum test. Results: The WU model to predict RP performed poorly for the RTOG 9311 data. The most predictive model in the RTOG 9311 dataset was a single-parameter model, D15 (r = 0.28). Combining the datasets, the best derived model was a two-parameter model consisting of mean lung dose and superior-to-inferior gross tumor volume position (r = 0.303). An equation and nomogram to predict the probability of RP was derived using the combined patient population. Conclusions: Statistical models derived from a large pool of candidate models resulted in well-tuned models for each subset (WU or RTOG 9311), which did not perform well when applied to the other dataset. However, when the data were combined, a model was generated that performed well on each data subset. The final model incorporates two effects: greater risk due to inferior lung irradiation, and greater risk for increasing normal lung mean dose. This formula and nomogram may aid clinicians during radiation treatment planning for lung cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

2021 Smoky Mountains Conference Data Challenge Synthetic-to-Real Domain Adaptation for Autonomous Driving Dataset

The dataset is comprised of both real and synthetic images from a vehicle's forward-facing camera. Each camera image is accompanied by a corresponding pixel-level semantic segmentation image (all files are .png files). In total, the dataset contains 5600 images in the training/validation set and 1400 images in the testing set. The training dataset contains mostly synthetic RGB images collected with a wide range of weather and lighting conditions using the CARLA simulator [1]. In addition, the training data also includes a small pre-selected subset of data from the Cityscapes training dataset – which is comprised of RGB-segmentation image pairs from driving scenarios in various European cities [2]. The testing data is split into three sets. The first set contains synthetic CARLA images with weather/lighting conditions that were not present in the training set. The second set is a subset of the Cityscapes testing dataset. Finally, the third set is an unknown testing set which will not be revealed to the participants until after the submission deadline. [1] Dosovitskiy, A., Ros, G., Codevilla, F., Lopez, A., and Koltun, V. (2017, October). CARLA: An open urban driving simulator. In Conference on robot learning (pp. 1-16). PMLR. [2] Cordts, M., Omran, M., Ramos, S., Rehfeld, T., Enzweiler, M., Benenson, R., ... and Schiele, B. (2016). The cityscapes dataset for semantic urban scene understanding. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 3213-3223).

99 GENERAL AND MISCELLANEOUS↗

Modeling the Risk of Radiation-Induced Acute Esophagitis for Combined Washington University and RTOG Trial 93-11 Lung Cancer Patients

Purpose: To construct a maximally predictive model of the risk of severe acute esophagitis (AE) for patients who receive definitive radiation therapy (RT) for non-small-cell lung cancer. Methods and Materials: The dataset includes Washington University and RTOG 93-11 clinical trial data (events/patients: 120/374, WUSTL = 101/237, RTOG9311 = 19/137). Statistical model building was performed based on dosimetric and clinical parameters (patient age, sex, weight loss, pretreatment chemotherapy, concurrent chemotherapy, fraction size). A wide range of dose-volume parameters were extracted from dearchived treatment plans, including Dx, Vx, MOHx (mean of hottest x% volume), MOCx (mean of coldest x% volume), and gEUD (generalized equivalent uniform dose) values. Results: The most significant single parameters for predicting acute esophagitis (RTOG Grade 2 or greater) were MOH85, mean esophagus dose (MED), and V30. A superior-inferior weighted dose-center position was derived but not found to be significant. Fraction size was found to be significant on univariate logistic analysis (Spearman R = 0.421, p < 0.00001) but not multivariate logistic modeling. Cross-validation model building was used to determine that an optimal model size needed only two parameters (MOH85 and concurrent chemotherapy, robustly selected on bootstrap model-rebuilding). Mean esophagus dose (MED) is preferred instead of MOH85, as it gives nearly the same statistical performance and is easier to compute. AE risk is given as a logistic function of (0.0688 Asterisk-Operator MED+1.50 Asterisk-Operator ConChemo-3.13), where MED is in Gy and ConChemo is either 1 (yes) if concurrent chemotherapy was given, or 0 (no). This model correlates to the observed risk of AE with a Spearman coefficient of 0.629 (p < 0.000001). Conclusions: Multivariate statistical model building with cross-validation suggests that a two-variable logistic model based on mean dose and the use of concurrent chemotherapy robustly predicts acute esophagitis risk in combined-data WUSTL and RTOG 93-11 trial datasets.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Machine-learning informed prediction of high-entropy solid solution formation: Beyond the Hume-Rothery rules

The empirical rules for the prediction of solid solution formation proposed so far in the literature usually have very compromised predictability. Some rules with seemingly good predictability were, however, tested using small data sets. Based on an unprecedented large dataset containing 1252 multicomponent alloys, machine-learning methods showed that the formation of solid solutions can be very accurately predicted (93%). The machine-learning results help identify the most important features, such as molar volume, bulk modulus, and melting temperature. As such a new thermodynamics-based rule was developed to predict solid–solution alloys. The new rule is nonetheless slightly less accurate (73%) but has roots in the physical nature of the problem. The new rule is employed to predict solid solutions existing in the three blocks, each of which consists of 9 elements. The predictions encompass face-centered cubic (FCC), body-centered cubic (BCC), and hexagonal closest packed (HCP) structures in a high throughput manner. The validity of the prediction is further confirmed by CALculations of PHAse Diagram (CALPHAD) calculations with high consistency (94%). Since the new thermodynamics-based rule employs only elemental properties, applicability in screening for solid solution high-entropy alloys is straightforward and efficient.

36 MATERIALS SCIENCE↗