SEARCH · Engineering Papers
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.
The raw bulk metagenomic data were trimmed and de-novo assembled. The assemblies were screened for viral contigs (by virsorter2, Deepvirfinder, VIBRANT, checkV) and binned to generate MAGs that were further refined by Hi-C read links.
During the development of high-consequence items, test systems should be capable of differentiating between test failures resulting from narrowly missing requirements versus those indicating potentially catastrophic faults. In many instances, classifying the data corresponds to simply identifying whether measured waveforms have approximately the anticipated shape. Cast in this light, the problem reduces to converting raw data into a form optimal for use with neural network classifiers. This manuscript investigates different means of representing raw data for image classification. Raw data plots and Short Time Fourier Transform (STFT) spectrograms are classified by both custom built, small-scale, Convolution Neural Networks (CNN) and open-source, multi-million parameter, pre-trained deep CNNs. In the case of time varying frequency content, the STFTs provide images with greater detail and can be accurately classified with simpler networks. This requires less memory and runs faster than classifying the raw data using the more sophisticated options—making STFTs optimal for applications with memory constraints. STFTs are not a panacea. In some cases the time-domain signal contains useful information that should not be discarded. Rather than using raw data or STFTs, the images can be constructed from both by using red and green channels of an RGB image to visualize the real and imaginary components of the transform, with the raw data occupying the blue channel.
Atmospheric water vapor pressure is an essential meteorological control on land surface and hydrologic processes. It is not as frequently observed as other meteorologic conditions, but often inferred through the August–Roche–Magnus formula by simply assuming dew point and daily minimum temperatures are equivalent or by empirically correlating the two temperatures using an aridity correction. The performance of both methods varies considerably across different regions and during different time periods; obtaining consistently accurate estimates across space and time remains a great challenge. We applied an interpretable Long Short-Term Memory (iLSTM) network conditioned on static, location specific attributes to estimate daily vapor pressure for 83 FLUXNET sites in the United States and Canada. This data package includes all raw data of the 83 FLUXNET sites, input data for model training/validation/test, trained models and results, and python codes for the manuscript "Improving the Estimation of the Atmospheric Water Vapor Pressure Using an Interpretable Long Short-term Memory Network". Specifically, it consists of five parts. - First, "1_Daymet_data_83sites.zip" includes raw data downloaded from Daymet for the 83 sites used in the paper according to their longitude and latitude, in which vapor pressure is used. It also includes a pre-processed CSV data file combining all data from the 83 sites which is specifically used for the paper. - Second, "2_Fluxnet2015_data_83sites.zip" includes raw half hourly data of the 83 sites downloaded from FLUXNET2015 data portal, pre-processed daily data of the 83 sites, a CSV file including combined pre-processed daily data of the 83 sites, and a CSV file including the information (site ID, site name, latitude, longitude, data available period) of the 83 sites. - Third, "3_MODIS_LAI_data_83sites_raw.zip" includes raw leaf area index (LAI) data downloaded from the AppEEARs data portal. - Fourth, "4_Scripts.zip" includes all scripts related to model training and post-processing of a trained model, and a jupyter notebook showing an example for model post-processing. Two typo errors in files titled "run2get_args.py" and "postprocess.py" were corrected on March 27, 2024 to avoid confusions. - Finally, "Trained_models_and_results.zip" includes three folders and three files with suffix ".npy", and each folder corresponds to one file with suffix ".npy" with the same title. Each of the three folders include all trained models associated with one iLSTM model configuration (35 models for each configuration, details are described in the paper). Each file with suffix ".npy" includes the post-processed results of the corresponding 35 models under one iLSTM model configuration.
Abstract Machine learning (ML) methods applied in scientific research often deal with interrelated features in high‐dimensional data. Reducing data noise and redundancy is needed to increase prediction accuracy and efficiency especially when dealing with data from field sensors. We explored an unsupervised learning method, the denoising autoencoder (DAE), to extract the underlying data structure from noisy raw data in the context of predicting hydrologic quantities from multiple field sensors. These sensors have intrinsic instrumental noise and occasional malfunctions that cause missing values. Our DAE neural network reconstructed meteorological sensor data containing noise and missing values to predict evapotranspiration in a mountainous watershed. The DAE reconstructed the sensor variables with a mean coefficient of determination value of 0.77 across 15 dimensions representing individual sensors. It reduced variance and bias uncertainties compared to a classical autoencoder model. The reconstruction quality varied across dimensions depending on their cross‐correlation and alignment with the underlying data structure. Uncertainties arising from the model structure were overall higher than those resulting from data corruption. We attached the DAE structure to a downstream ET‐prediction neural network in three formats and achieved reasonably accurate ET predictions . The use of the DAE notably reduced variance uncertainty in ET prediction. However, excessive variance reduction may be accompanied by an increase in bias due to the intrinsic bias‐variance tradeoff. Our method of evaluating and reducing uncertainties in aggregated data from different sources can be used to improve predictive models, process understanding, and uncertainty quantification for better water resource management. Plain Language Summary We present a machine learning method, namely the denoising autoencoder, which reduces the effects of data noise and missing values typically present in scientific data sets collected through sensor measurements. This method selects the most relevant information from noisy raw data collected by the instruments and fills in missing values. To demonstrate the effectiveness of our method, we applied it to predict evapotranspiration, a hydrologic variable that represents the water moved from the land surface to the atmosphere through a combination of evaporation and plant water use (transpiration). We also used a random sampling technique (the Monte Carlo method) to compare the uncertainty in the predictions when using the raw and noisy data versus the reconstructed data. The denoising process produced more accurate predictions of evapotranspiration with less uncertainty. Improved predictions of evapotranspiration can lead to a better understanding and accounting of water budgets. This ML approach is broadly suitable for a wide variety of applications that involve noisy sensor data with missing values. Key Points We used a denoising autoencoder (DAE) neural network to reduce noise in meteorological and soil sensor observations by on average We used Monte Carlo sampling to estimate the bias and variance of all model outputs, including uncertainty sources from data and the model We attached the DAE component to a downstream neural network to predict ET with the variance reduced by , compared to that without the DAE
As the Tank Operations Contractor for U.S. Department of Energy operations at the Hanford site in Washington State, Washington River Protection Solutions (WRPS) is responsible for managing highly radioactive wastes stored in tanks at Hanford. To protect workers at Hanford Tank Farms, WRPS tests air-purifying respirator (APR) and powered air-purifying respirator (PAPR) chemical cartridges commonly used at the tank farms. The tests were conducted to determine the period of time the cartridges would provide adequate performance for APRs and PAPRs when workers are exposed to a mixture of Chemicals of Potential Concern (COPC) from any vapors exiting headspaces in the storage tanks. Occupational Safety and Health Administration (OSHA) Standard 29 Code of the Federal Regulations (CFR) 1910.134(d)(3)(iii)(b)(2) specifies that for protection against gases and vapors, employers shall implement a schedule for cartridges to ensure that change-outs occur before the end of service life. The change schedule can be based on objective information or data that ensures cartridge change-outs occur before the end of their service life. The primary function of the WRPS Cartridge Test Program is to obtain objective data to determine service lives for the APR and PAPR cartridges used at Hanford Tank Farms. WRPS contracted with Pacific Northwest National Laboratory to analyze the test data and offer an independent analysis and any recommendations. This report summarizes data analysis of APR and PAPR cartridge testing on a vapor slipstream from the Hanford AX tank farm exhauster. Volume 1 of this report documents the testing, data analysis, results, conclusions, and recommendations resulting from the cartridge testing on AX exhauster slipstream vapors. Volume 2 provides an introduction to the raw data, including analytical laboratory analysis results that supported the analysis and conclusions documented in Volume 1.
As the Tank Operations Contractor for U.S. Department of Energy operations at the Hanford site in Washington State, Washington River Protection Solutions (WRPS) is responsible for managing highly radioactive wastes stored in tanks at Hanford. To protect workers at Hanford Tank Farms, WRPS tests air-purifying respirator (APR) and powered air-purifying respirator (PAPR) chemical cartridges commonly used at the tank farms. The tests were conducted to determine the period of time the cartridges would provide adequate performance for APRs and PAPRs when workers are exposed to a mixture of Chemicals of Potential Concern (COPC) from any vapors exiting headspaces in the storage tanks. Occupational Safety and Health Administration (OSHA) Standard 29 Code of the Federal Regulations (CFR) 1910.134(d)(3)(iii)(b)(2) specifies that for protection against gases and vapors, employers shall implement a schedule for cartridges to ensure that change-outs occur before the end of service life. The change schedule can be based on objective information or data that ensures cartridge change-outs occur before the end of their service life. The primary function of the WRPS Cartridge Test Program is to obtain objective data to determine service lives for the APR and PAPR cartridges used at Hanford Tank Farms. WRPS contracted with Pacific Northwest National Laboratory to analyze the test data and offer an independent analysis and any recommendations. This report summarizes data analysis of APR cartridge testing on headspace vapors from Hanford SX-101 and SX-104 single-shell tanks. Volume 1 of this report documents the testing, data analysis, results, conclusions, and recommendations resulting from the APR testing on SX-101 and SX-104 headspace vapors. Volume 2 provides an introduction to the raw data, including analytical laboratory analysis results that supported the analysis and conclusions documented in Volume 1.
As the Tank Operations Contractor for U.S. Department of Energy operations at the Hanford site in Washington State, Washington River Protection Solutions (WRPS) is responsible for managing highly radioactive wastes stored in tanks at Hanford. WRPS tests of cartridges used in air-purifying respirators (APR) and powered air-purifying respirators (PAPR) commonly used at Hanford Tank Farms. The tests are conducted to determine the period of time the cartridges would provide adequate performance for APRs and PAPRs used to protect workers when exposed to a mixture of Chemicals of Potential Concern (COPC) from any vapors exiting headspaces in the storage tanks. Occupational Safety and Health Administration (OSHA) Standard 29 Code of the Federal Regulations (CFR) 1910.134(d)(3)(iii)(b)(2) specifies that for protection against gases and vapors, employers shall implement a schedule for cartridges to ensure that change-outs occur before the end of service life. The change schedule can be based on objective information or data that ensures cartridge change-outs occur before the end of their service life. The primary function of the WRPS Cartridge Test Program is to obtain objective data to determine service lives for the APR and PAPR cartridges used at Hanford Tank Farms. WRPS contracted with Pacific Northwest National Laboratory to analyze the test data and offer an independent analysis and any recommendations. This report summarizes data analysis of PAPR cartridge testing on headspace vapors from Hanford SX-101 and SX-104 single-shell tanks. Two different PAPR cartridges—one from MSA Safety Inc. (Pittsburgh, Pennsylvania) and another from 3M (Maplewood, Minnesota)—were assessed on each tank headspace source on separate days. These data represent the first PAPR cartridge testing under the recent WRPS program, as testing to date had been focused on APR cartridges. Volume 1 of this report documents the testing, data analysis, results, conclusions, and recommendations resulting from the PAPR testing on SX-101 and SX-104 headspace vapors. Volume 2 provides an introduction to the raw data, including analytical laboratory analysis results that supported the analysis and conclusions documented in Volume 1.
As the Tank Operations Contractor for U.S. Department of Energy operations at the Hanford site in Washington State, Washington River Protection Solutions (WRPS) is responsible for managing highly radioactive wastes stored in tanks at Hanford. WRPS tests air-purifying respirator (APR) and powered air-purifying respirator (PAPR) chemical cartridges commonly used by workers at Hanford Tank Farms to determine the period of time that the cartridges would provide adequate performance for APRs and PAPRs used to protect workers when exposed to a mixture of Chemicals of Potential Concern (COPC) from any vapors exiting headspaces in the storage tanks. The Occupational Safety and Health Administration (OSHA) Standard promulgated in Title 29 of the Code of the Federal Regulations (CFR) 1910.134(d)(3)(iii)(b)(2) specifies that for protection against gases and vapors, employers shall implement a schedule for cartridges to ensure that change-outs occur before the end of service life. The change schedule can be based on objective information or data that ensures cartridge change-outs occur before the end of their service life. The primary function of the WRPS Cartridge Test Program is to obtain objective data to determine service lives for the APR and PAPR cartridges used at Hanford Tank Farms. WRPS contracted with Pacific Northwest National Laboratory to analyze the test data and offer an independent analysis and any recommendations. This report summarizes data analysis of APR and PAPR cartridge testing on BY-110 and BY-108 headspace vapors, respectively. Two different APR cartridges from SCOTT Safety (Monroe, North Carolina) were assessed for the BY-110 headspace vapors, and two different PAPR cartridges—one from MSA Safety Inc. (Pittsburgh, Pennsylvania) and another from 3M (Maplewood, Minnesota)—were assessed for the BY-108 headspace vapors. Volume 1 of this report documents the testing, data analysis, results, conclusions, and recommendations resulting from the cartridge testing of vapors from the BY-110 and BY-108 headspaces. Volume 2 provides an introduction to the raw data, including analytical laboratory analysis results that supported the analysis and conclusions documented in Volume 1.
These files contain the geodatabases related to the Desert Peak Geothermal Field. It includes all input and output files used in the project. The files include data categories of raw data, pre-processed data, and analysis (post-processed data). In each of these categories there are six additional types of raster catalogs including Radar, SWIR, Thermal, Geophysics, Geology, and Wells. The files for the Desert Peak Geothermal Site are used with the Geothermal Exploration Artificial Intelligence to identify indicators of blind geothermal systems. The included zip file is a geodatabase to be used with ArcGIS and the tar file is an inclusive database that encompasses the inputs and outputs for the Desert Peak Geothermal Field.
This submission includes processed and reduced data for circulation testing that was conducted at the 164' fracture on the 4850 ft level of the Sanford Underground Research Facility. The circulation tests were done to test the flow through the 164' fracture in the EGS Collab Experiment 1 testbed on the 4850 ft level of the Sanford Underground Research Facility in order to validate computer models concerned with flow processes and heat exchange processes in EGS. This data was processed from raw data that is available in the GDR submission linked below: "EGS Collab Circulation Testing Raw Data and Documentation." Python scripts used to process and reduce the data are included. The scripts also serve as "documentation" as well as examples for data processing. More information about the test itself as well as the rationale and process of data processing is available at wiki page linked below.
As the Tank Operations Contractor for U.S. Department of Energy operations at the Hanford site in Washington State, Washington River Protection Solutions (WRPS) is responsible for managing highly radioactive wastes stored in tanks at Hanford. WRPS tests air-purifying respirator (APR) and powered air-purifying respirator (PAPR) chemical cartridges commonly used at Hanford Tank Farms to determine the period of time that the cartridges would provide adequate performance for APRs and PAPRs used to protect workers when exposed to a mixture of Chemicals of Potential Concern (COPC) from any vapors exiting headspaces in the storage tanks. The Occupational Safety and Health Administration (OSHA) Standard promulgated in Title 29 of the Code of the Federal Regulations (CFR) 1910.134(d)(3)(iii)(b)(2) specifies that for protection against gases and vapors, employers shall implement a schedule for cartridges to ensure that change-outs occur before the end of service life. The change schedule can be based on objective information or data that ensures cartridge change-outs occur before the end of their service life.[2-5] The primary function of the WRPS Cartridge Test Program is to obtain objective data to determine service lives for the APR and PAPR cartridges used at Hanford Tank Farms. WRPS contracted with Pacific Northwest National Laboratory to analyze the test data and offer an independent analysis and any recommendations. Volume 1 of this report summarizes data analysis of APR and PAPR cartridge testing on vapors from the AP tank farm exhauster. Previous testing of APR cartridges was conducted on the AP exhauster in June 2016. However, an AP exhauster upgrade was completed in September 2016. Two different APR cartridges from SCOTT Safety (Monroe, North Carolina) were assessed for the new AP exhauster source, along with two different PAPR cartridges—one from MSA Safety Inc. (Pittsburgh, Pennsylvania) and another from 3M (Maplewood, Minnesota). Volume 2 provides an introduction to the raw data, including analytical laboratory analysis results that supported the analysis and conclusions documented in Volume 1.
An integrated open mineral spectral library designed to enhance the utility and precision of mineral spectral data for geothermal exploration, developed from a reliable and comprehensive digital dataset for seamless sharing by integrating field data, the USGS spectral library, and pertinent information obtained from Coso geothermal field (Coso) in California. The ASD FieldSpec portable spectrometer was utilized for collecting spectral data, which was subsequently analyzed using the THOR Material Identification tool in ENVI, The Spectral Geologist (TSG) software by CSIRO, and the Fully Constrained Linear Spectral Unmixing algorithm (FCLSU) in MATLAB. Scanning Electron Microscopy (SEM) with a mineralogy-analyzing function was employed to assess the mineral composition of samples, ensuring precise mineralogical analysis. A portable X-ray fluorescence (pXRF) spectrometer was also utilized to gather information on elemental enrichment. A framework for developing spectra data and establishing spectral libraries for various geological cases was proposed within this study. The characteristic spectra of six alteration minerals - alunite, chalcedony, epidote, hematite, kaolinite, and opal - were acquired from Coso samples. The spectral library for the Coso alteration minerals was introduced for further application in academic study or industrial exploration. To browse the Coso Geothermal Spectral data and related figures from spreadsheets: #1 Unzip and store the following items in the same folder. 'Contact Probe Data.zip', 'Sample Photos.zip', and 'Coso spectra of higher-certainity minerals.xlsx'. #2 Open 'Coso spectra of higher-certainity minerals.xlsx'. The hyperlinks in the spreadsheet lead to the folders or figures of: spectra .asd file, spectra ASC II file, spectra plots, and sample photos. The spectra data is raw data without splice correction. Spectra .asd files require particular software to open. (These cannot be opened in GIS software such as ArcGIS.) Spectra ASC II files can be opened in a text editor or spread sheet program.
This dataset includes modeled velocity and discharge at five communities in the middle Kuskokwim River region: Aniak, Chuathbaluk, Crooked Creek, Red Devil and Stony River. Modeled velocities and discharge represent daily averages calculated for the openwater season (OWS) from June 1 - October 18 over the 20 year period 2000-2019 using the raw data described below and included in this archive; full details of methodology are described in (Brown et al. submitted to Renewable Energy). Raw data inputs to inform the modeling process include in-situ measurements of 1) discharge with an acoustic Doppler current meter (ADCP, 600kHz Workhorse Rio Grande by Teledyne RD Instruments) and a global positioning receiver (GPS, Trimble 5700, 5800 and R8) utilizing Real Time Kinematic (RTK) GPS mode over 1-2 days at each site in 2009 or 2010 (Ravens 2014), and 2) river stage with a water level logger (HOBO U20-001-01 by Onset) over 2-9 weeks at each site (Ravens 2014), 3) in addition to a 20 year long-term discharge record collected at the USGS stream gage site in Crooked Creek (USGS 2016). Raw data (discharge and stage) are included in this archive for two additional communities: Lower Kalskag and Sleetmute, where modeled velocities were not calculated due to equipment failure or loss. The USGS stream gage data at Crooked Creek (USGS 2016) and stream gage methodology (Turnipseed and Sauer 2010) are publicly available online, so the data are not duplicated here.
The following submission includes raw and processed data from the in water deployment of NREL's Hydraulic and Electric Reverse Osmosis Wave Energy Converter (HERO WEC), in the form of parquet files, TDMS files, CSV files, bag files and MATLAB workspaces. This dataset was collected in March 2024 at the Jennette's pier test site in North Carolina. This submission includes the following: - Data description document (HERO WEC FY24 Hydraulic Deployment Data Descriptions.doc) - This document includes detailed descriptions of the type of data and how it was processed and/or calculated. - Processed MATLAB workspace - The processed data is provided in the form of a single MATLAB workspace containing data from the full deployment. This workspace contains data from all sensors down sampled to 10 Hz along with all array Value Added Products (VAPs). - MATLAB visualization scripts - The MATLAB workspaces can be visualized using the file "HERO_WEC_2024_Hydraulic_Config_Data_Viewer.m/mlx". The user simply needs to download the processed MATLAB workspaces, specify the desired start and end times and run this file. Both the .m and .mlx file format has been provided depending on the user's preference. - Summary Data - The fully processed data was used to create a summary data set with averages and important calculations performed on 30-minute intervals to align with the intervals of wave resource data reported from nearby CDIP ocean observing buoys located 20km East of Jennette's pier and 40km Northeast of Jennette's pier. The wave resource data provided in this data set is to be used for reference only due the difference in water depth and proximity to shore between the Jennette's pier test site and the locations of the ocean observing buoys. This data is provided in the Summary Data zip folder, which includes this data set in the form of a MATLAB workspace, parquet file, and excel spreadsheet. - Processed Parquet File - The processed data is provided in the form of a single parquet file containing data from all HERO WEC sensors collected during the full deployment. Data in these files has been down sampled to 10 Hz and all array VAPs are included. - Interim Filtered Data - Raw data from each sensor group partitioned into 30-minute parquet files. These files are outputs from an intermediate stage of data processing and contain the raw data with no Quality Control (QC) or calculations performed in a format that is easier to use than the raw data. - Raw Data - Raw, unprocessed data from this deployment can be found in the Raw Data zip folder. This data is provided in the form of TDMS, CSV, and bag files in the original format output by the MODAQ system. - Python Data Processing Script - This links to an NREL public github repository containing the python script used to go from raw data to fully processed parquet files. Additional documentation on how to use this script is included in the github repository. This data set has been developed by the National Renewable Energy Laboratory, operated by Alliance for Sustainable Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. Funding provided by the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Water Power Technologies Office.
A central repository of thermophysical and thermochemical properties of molten salt compositions of relevance to molten salt reactors (MSRs) is vital in supporting the broad community of MSR developers, who are at various stages of developing and deploying their reactor designs. In general, these MSR designs differ significantly from developer to developer (e.g., with respect to the hardness of the neutron spectra, level of fissile loading, target multicomponent temperatures and power levels, and moderating capabilities). Therefore, the fuel and coolant salts being considered vary greatly: they may be chlorides or fluorides, they utilize different actinides at different ratios, and the cations in the melt are selected based on perceived advantages and disadvantages. Considering the general need for thermal properties, and the vastness of the array of potential candidate salt mixtures, the Molten Salt Thermal Properties Database (MSTDB) was initiated in 2018 with the goal of providing thermophysical and thermochemical characterization of key molten salt compounds and mixtures across their temperature and compositional domains. The MSTDB is thus divided into the thermophysical arm (MSTDB-TP) and the thermochemical arm (MSTDB-TC). The MSTDB is an effort funded by the Department of Energy, Office of Nuclear Energy (DOE-NE) Nuclear Energy Advanced Modeling and Simulation (NEAMS) program, and the MSR Campaign. This report provides an overview of the MSTDB-TP v4.0 in terms of the data contained within, the state of the tools used to access the data, the availability of predictive models that leverage the raw data in the database, the preliminary status of developmental efforts that are currently underway, and an account of future goals for MSTDB-TP. The primary goal for the update from MSTDB-TP v.3.1 to v4.0 was the incorporation of surface tension data into the database; this property is important for thermal hydraulics modeling and species transport in other tools that have been developed under the NEAMS program. A breakdown of the surface tension data that have been added into MSTDB-TP v4.0 is provided herein, and the manner in which the quality of the data has been assessed is also documented. For MSTDB-TP v4.0, newly published thermophysical property data—primarily from collaborative experimental efforts under the MSR Campaign—have been incorporated into the database, and the resulting expansion is documented here. Because of the size to which MSTDB-TP has grown, the raw data format has now been recast into JavaScript Object Notation (JSON) format for easier connection with the MSTDB-TP application programming interface (API). Saline; the pre-existing comma-separated value (CSV) format has been deprecated but is still maintained, accessible, and up to date. As a final effort in packaging the MSTDB-TP v4.0 update, the graphical user interface (GUI) for MSTDB has been updated to allow full accessibility to the density and viscosity predictive models, which are based on Redlich-Kister expansions of MSTDB-TP raw data. Some other major aspects of this report, in terms of preliminary and future work, include: (1) documentation of the formalism and preliminary testing of a kinetic theory model that may act as a predictive model for thermal conductivity; (2) documentation of the candidate predictive models that may be considered in the future for surface tension, making use of the surface tension data now in MSTDB-TP v4.0; (3) a preliminary account of a data collection process that will enable the filling of additional gaps within MSTDB-TP, namely with data which have been collected computationally (e.g., through ab initio molecular dynamics).
Raw Mass Spectrometry data generated for the manuscript "Rubisco kinetic acclimation at the holoenzyme level." It characterizes the stoichiometry of Rubisco subunits in different Rubisco samples.
Raw data for Figure 2 is located in: /Experiments/Fig2_delayed_aggregation Raw data for Figures 3 and 4 can be found at: /Experiments/Fig3_4_Core_shell The analysis code is available at: /Experiments/Code Raw data for the Figure 5 simulation is locate in: /Simulation/Raw_data/Fig5_Structural_features_of_cluster
There is a growing body of work that leverages features extracted via topological data analysis to train machine learning models. While this field, sometimes known as topological machine learning (TML), has seen some notable successes, an understanding of how the process of learning from topological features differs from the process of learning from raw data is still limited. In this work, we begin to address one component of this larger issue by asking whether a model trained with topological features learns internal representations of data that are fundamentally different than those learned by a model trained with the original raw data. To quantify ``different'', we exploit two popular metrics that can be used to measure the similarity of the hidden representations of data within neural networks, neural stitching and centered kernel alignment. From these we draw a range of conclusions about how training with topological features does and does not change the representations that a model learns. Perhaps unsurprisingly, we find that structurally, the hidden representations of models trained and evaluated on topological features differ substantially compared to those trained and evaluated on the corresponding raw data. On the other hand, our experiments show that in some cases, these representations can be reconciled (at least to the degree required to solve the corresponding task) using a simple affine transformation. We conjecture that this means that neural networks trained on raw data may extract some limited topological features in the process of making predictions.