Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “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 1,621 records · Page 90

Sampling Subjective Polygons for Patch-Based Deep Learning Land-Use Classification in Satellite Images

Model generalization remains a key challenge in the analysis of large amounts of heterogeneous satellite image data. One major limiting factor in developing generalizable models, in the context of supervised learning, is the lack of high quality training datasets. A model's capacity to perform well on new data is often inhibited by imbalance and bias in the data that was used for training. This is especially a problem when using convolutional neural networks to classify urban land-use in satellite images. Notable dataset imbalance issues in this application include land-use type imbalance and image scene imbalance. To begin understanding these dataset imbalance problems in more detail, we develop and test a number of sampling methods for generating training image datasets from subjective training polygons for urban land-use classification. We investigate sampling at different point densities as a means to reduce content repetition and therefore content imbalance and bias in the training image dataset.

Arndt, Jacob↗

Domain Adaptation for Convolutional Neural Networks Based Remote Sensing Scene Classification

Remote sensing (RS) scene classification plays an important role in the field of earth observation. With the rapid development of the remote sensing techniques, a large number of RS scene images are available. Since manually labeling largescale RS scene images is both labor and time consuming, when a new unlabeled dataset is obtained, how to use the existing labeled datasets to classify the new unlabeled images is an important research direction. Different RS scene image datasets may be taken from different type of sensors, and the images may vary from imaging modalities, spatial resolutions and image scales, so the distribution discrepancy exists among different image datasets. As a result, simply applying Convolutional Neural Networks (CNN) trained on source domain cannot accurately classify the images on target domain. Domain Adaptation (DA) can be helpful to solve this problem. In this paper, we design a Subspace Alignment (SA) and CNN based framework to solve the DA problem in RS scene image classification. A new SA layer is proposed and added into CNN models for DA, which could align the source and target domains in some feature subspace. Finetuning the modified CNN model with the added SA layer is able to make the CNN model adapt to the aligned feature subspace and helps to relieve the domain distribution discrepancy. The experiments conducted on two public datasets show that adding the SA layer into CNN can improve the scene classification on target domain.

97 MATHEMATICS AND COMPUTING↗

A data-driven approach to sampling matrix selection for compressive sensing

Sampling is a fundamental aspect of any implementation of compressive sensing. Typically, the choice of sampling method is guided by the reconstruction basis. However, this approach can be problematic with respect to certain hardware constraints and is not responsive to domain-specific context. We propose a method for defining an order for a sampling basis that is optimal with respect to capturing variance in data, thus allowing for meaningful sensing at any desired level of compression. We focus on the Walsh-Hadamard sampling basis for its relevance to hardware constraints, but our approach applies to any sampling basis of interest. We illustrate the effectiveness of our method on the Physical Sciences Inc. Fabry-Perot interferometer sensor multispectral dataset, the Johns Hopkins Applied Physics Lab FTIR-based longwave infrared sensor hyperspectral dataset, and a Colorado State University Swiss Ranger depth image dataset. The spectral datasets consist of simulant experiments, including releases of chemicals such as GAA and SF6. We combine our sampling and reconstruction with the adaptive coherence estimator (ACE) and bulk coherence for chemical detection and we incorporate an algorithmic threshold for ACE values to determine the presence or absence of a chemical. We compare results across sampling methods in this context. We have successful chemical detection at a compression rate of 90%. For all three datasets, we compare our sampling approach to standard orderings of sampling basis such as random, sequency, and an analog of sequency that we term `frequency.' In one instance, the peak signal to noise ratio was improved by over 30% across a test set of depth images.

compressive sensing, hyperspectral imaging, Walsh-↗

Multiattribute probabilistic prostate elastic registration (MAPPER): Application to fusion of ultrasound and magnetic resonance imaging

Purpose: Transrectal ultrasound (TRUS)-guided needle biopsy is the current gold standard for prostate cancer diagnosis. However, up to 40% of prostate cancer lesions appears isoechoic on TRUS. Hence, TRUS-guided biopsy has a high false negative rate for prostate cancer diagnosis. Magnetic resonance imaging (MRI) is better able to distinguish prostate cancer from benign tissue. However, MRI-guided biopsy requires special equipment and training and a longer procedure time. MRI-TRUS fusion, where MRI is acquired preoperatively and then aligned to TRUS, allows for advantages of both modalities to be leveraged during biopsy. MRI-TRUS-guided biopsy increases the yield of cancer positive biopsies. In this work, the authors present multiattribute probabilistic postate elastic registration (MAPPER) to align prostate MRI and TRUS imagery. Methods: MAPPER involves (1) segmenting the prostate on MRI, (2) calculating a multiattribute probabilistic map of prostate location on TRUS, and (3) maximizing overlap between the prostate segmentation on MRI and the multiattribute probabilistic map on TRUS, thereby driving registration of MRI onto TRUS. MAPPER represents a significant advancement over the current state-of-the-art as it requires no user interaction during the biopsy procedure by leveraging texture and spatial information to determine the prostate location on TRUS. Although MAPPER requires manual interaction to segment the prostate on MRI, this step is performed prior to biopsy and will not substantially increase biopsy procedure time. Results: MAPPER was evaluated on 13 patient studies from two independent datasets—Dataset 1 has 6 studies acquired with a side-firing TRUS probe and a 1.5 T pelvic phased-array coil MRI; Dataset 2 has 7 studies acquired with a volumetric end-firing TRUS probe and a 3.0 T endorectal coil MRI. MAPPER has a root-mean-square error (RMSE) for expert selected fiducials of 3.36 ± 1.10 mm for Dataset 1 and 3.14 ± 0.75 mm for Dataset 2. State-of-the-art MRI-TRUS fusion methods report RMSE of 3.06–2.07 mm. Conclusions: MAPPER aligns MRI and TRUS imagery without manual intervention ensuring efficient, reproducible registration. MAPPER has a similar RMSE to state-of-the-art methods that require manual intervention.

60 APPLIED LIFE SCIENCES↗

Novelty and uniqueness patterns of rare members of the soil biosphere

Soil bacterial communities typically exhibit a distribution pattern in which most bacterial species are present in low abundance. Due to the relatively small size of most culture independent sequencing surveys, a detailed phylogenetic analysis of rare members of the community is lacking. To gain access to the rarely sampled soil biosphere, we analyzed a dataset of 13,001 near full-length 16S rRNA gene clones derived from an undisturbed tall grass prairie soil in central Oklahoma. Rare members of the soil bacterial community (empirically defined at two different abundance cutoffs) represented 18.1%-37.1% of the total number of clones in the dataset and were, on average, less similar to their closest relatives in public databases when compared to more abundant members of the community. Detailed phylogenetic analyses indicated that members of the soil rare biosphere either belonged to novel bacterial lineages (members of five novel bacterial phyla identified in the dataset, as well as members of multiple novel lineages within previously described phyla or candidate phyla), lineages that are prevalent in other environments but rarely encountered in soil, or were close relatives to more abundant taxa in the dataset. While a fraction of the rare community was closely related to more abundant taxonomic groups in the dataset, a significant portion of the rare biosphere represented evolutionarily distinct lineages at various taxonomic cutoffs. We reason that recognizing these novelty and uniqueness patterns is key to understanding the origins, dynamics, and potential ecological roles of members of the soil’s rare biosphere.

54 ENVIRONMENTAL SCIENCES↗

Citations and Readership are Poor Indicators of Research Excellence

In this paper we show that citation counts and Mendeley readership are poor indicators of research excellence. Our experimental design builds on the assumption that a good evaluation metric should be able to distinguish publications that have changed a research field from those that have not. The experiment has been conducted on a new dataset for bibliometric research which we call TrueImpactDataset. TrueImpactDataset is a collection of research publications of two types -- research papers which are considered seminal work in their area and papers which provide a survey (a literature review) of a research area. The dataset also contains related metadata, which include DOIs, titles, authors and abstracts. We describe how the dataset was built and provide overview statistics of the dataset. We propose to use the dataset for validating research evaluation metrics. By using this data, we show that widely used research metrics only poorly distinguish excellent research.

Herrmannova, Dasha↗

Visual Analytics of Heterogeneous Data Using Hypergraph Learning

For real-world learning tasks (e.g., classification), graph-based models are commonly used to fuse the information distributed in diverse data sources, which can be heterogeneous, redundant, and incomplete. Thesemodels represent the relations in different datasets as pairwise links. However, these links cannot deal with high-order relations which connect multiple objects (e.g., in public health datasets, more than two patient groups admitted by the same hospital in 2014). In this article, we propose a visual analytics approach forthe classification on heterogeneous datasets using the hypergraph model. The hypergraph is an extension to traditional graphs in which a hyperedge connects multiple vertices instead of just two. We model various high-order relations in heterogeneous datasets as hyperedges and fuse different datasets with a unified hypergraph structure. We use the hypergraph learning algorithm for predicting missing labels in the datasets.To allow users to inject their domain knowledge into the model-learning process, we augment the traditional learning algorithm in a number of ways. In addition, we also propose a set of visualizations which enable the user to construct the hypergraph structure and the parameters of the learning model interactively during the analysis. Here, we demonstrate the capability of our approach via two real-world cases.

97 MATHEMATICS AND COMPUTING↗

Visual Instance-aware Prompt Tuning

Visual Prompt Tuning (VPT) has emerged as a parameter-efficient fine-tuning paradigm for vision transformers, with conventional approaches utilizing dataset-level prompts that remain the same across all input instances. We observe that this strategy results in sub-optimal performance due to high variance in downstream datasets. To address this challenge, we propose Visual Instance-aware Prompt Tuning (ViaPT), which generates instance-aware prompts based on each individual input and fuses them with dataset-level prompts, leveraging Principal Component Analysis (PCA) to retain important prompting information. Moreover, we reveal that VPT-Deep and VPT-Shallow represent two corner cases based on a conceptual understanding, in which they fail to effectively capture instance-specific information, while random dimension reduction on prompts only yields performance between the two extremes. Instead, ViaPT overcomes these limitations by balancing dataset-level and instance-level knowledge, while reducing the amount of learnable parameters compared to VPT-Deep. Extensive experiments across 34 diverse datasets demonstrate that our method consistently outperforms state-of-the-art baselines, establishing a new paradigm for analyzing and optimizing visual prompts for vision transformers.

Xiao, Xi [ORNL] (ORCID:0009000009316982)↗

High-Brightness Self-seeded X-ray Free Electron Laser to Precisely Map Macromolecular Structure

Two datasets were uploaded. Run numbers #r0101~#r0112 were collected from self-seeded XFEL mode. Run numbers #r0079~#r0088 and #r0090 were collected from SASE mode. The run number is shown in the file name as "ue_191027_SFX_runnumber-c00.cxi.". To understand the effect of self-seeded XFEL source, we used the datasets from run numbers #r0102 to #r0112. Datasets used were randomly chosen. - SS1 is from #r0102~#r0112. - SS2 is from #r0102~#r0105 and #r0109. - SS3 is from #r0104 and #r0105. To compare self-seeded mode and SASE mode, we also used corresponding three SASE datasets labeled as SASE1, SASE2, and SASE3 from run numbers #r0079 to #r0089. Datasets used were randomly chosen. - SASE1 is from #r0079~#r0089. - SASE2 is from #r0080, #r0081, #r0085, #r0087. - SASE3 is composed of #r0079, #r0080.

Lysozyme↗

stovertillage2019-bc1040-060

As part of the Billion Ton resource assessment projections created in 2016 (see https://www.energy.gov/sites/prod/files/2016/12/f34/2016_billion_ton_rep...), this dataset was produced and titled a "base-case" scenario. This broader dataset provided an updated assessment of the potential economic availability of biomass resources from agricultural lands reported at the farmgate under conservative assumptions. Crop residues quantified in this dataset include corn stover, cereal (wheat, oats, and barley) straws, and sorghum stubble. We have isolated corn stover in this dataset. What is the purpose of the data set? Why were the data collected?* Per request for use in subsequent research, we have isolated corn stover in 2019 from the broader base-case projections and have provided tillage classification details from this projection. Tillage classification assumptions in this scenario allow a moderate deviation from a baseline situation (using historic CTIC data on tillage type used in counties for each crop). This dataset allowed moderate flexibility of farmers to put land into another tillage type (no till, conservation till, and reduced till) where a higher net present value was calculated.

Davis, Maggie R.↗

Western Interconnect Transmission Line Wildfire Risk

The datasets contain wildfire risk probabilities for transmission lines in the Western Interconnect, identified only by the HIFLD dataset identifier. The dataset is generated using PNNL's WREST Tool (https://www.pnnl.gov/projects/wildfire-risk-resilience/tools/wrest). The input dataset includes HIFLD datasets on transmission lines and wildfire burn potentials and flame lengths from the PNNL RADR-FIRE Tool for 2025.

psps↗

Small Molecule Evaluator

Using machine learning to score potential drug candidates may offer an advantage over traditional imprecise scoring functions because the parameters and model structure can be learned from the data. However, models may lack interpretability, are often overfit to the data, and are not generalizable to drug targets and chemotypes not in the training data. Benchmark datasets are prone to artificial enrichment and analogue bias due to the overrepresentation of certain scaffolds in experimentally determined active sets. Datasets can be evaluated using spatial statistics to quantify the dataset topology and better understand potential biases. Dataset clumping comprises a combination of self-similarity of actives and separation from decoys in chemical space and is associated with overoptimistic virtual screening results. This code explores methods of quantifying potential biases and examines some common benchmark datasets.

Ellingson, Sally↗

Watershed Workflow

Watershed Workflow aims to lower the bar on leveraging open data products in hyperresolution watershed hydrologic models. Integrated, distributed hydrologic models of watersheds require large amounts and complexity of data including meteorological forcing datasets; land cover datasets; watershed geometry, hydrogography, and elevation datasets; and subsurface soil and rock structural datasets. Increasingly open data products meeting this need are available through U.S. governmental and other agencies via APIs and other electronic means, but are not available in the same format, through the same API, or in the same coordinate system as one another, and present other challenges in manipulations across scales. This product automates the discovery, download, curation, and integration of these types of datasets into a common format for use with hydrologic models.

Coon, EthanT [Oak Ridge National Laboratory] (0000↗

ODI notebook additive_manufacturing_video_2022

Additive Manufacturing, video dataset Two-photon lithography (TPL) is a widely used 3D nanoprinting technique that uses laser light to create objects. Challenges to large-scale adoption of this additive manufacturing method include identifying light dosage parameters and monitoring during fabrication. A research team from LLNL, Iowa State University, and Georgia Tech is applying machine learning models to tackle these challenges-i.e., accelerate the process of identifying optimal light dosage parameters and automate the detection of part quality. Funded by LLNL's Laboratory Directed Research and Development Program, the project team has curated a video dataset of TPL processes for parameters such as light dosages, photo-curable resins, and structures. Both raw and labeled versions of the datasets are available on the links in the Open Data Initiative page. The code uses the labeled dataset. Notebook compiled by Nisha Mulakken (mulakken1@llnl.gov) for LLNL Open Data Initiative, Summer 2022. Original code provided by research team. Publications: X.Y. Lee, S.K. Saha, S. Sarkar, B. Giera. "Automated detection of part quality during two-photon lithography via deep learning." Additive Manufacturing 36, December 2020: doi.org/10.1016/j.addma.2020.101444 X.Y. Lee, S.K. Saha, S. Sarkar, B. Giera. "wo Photon lithography additive manufacturing: Video dataset of parameter sweep of light dosages, photo-curable resins, and structures." Data in Brief 32, October 2020. doi.org/10.1016/j.dib.2020.106119.

Mulakken, NishaJ↗

gcm_eval (Global Climate Model Evaluation) [SWR-24-37]

The interplay between energy, climate, and weather is becoming more complex as our changing climate continues to affect the weather we experience which in turn drives changes in the ever increasing share of renewable energy generation and energy demand. Energy system analyses commonly rely on meteorological inputs to estimate renewable energy generation and energy demand; however, these inputs rarely represent the estimated impacts of future climate change. Climate models and publicly available climate change datasets can be used for this purpose, but the selection of inputs from the myriad of available models and datasets is a nuanced and often subjective process. There is no single perfect climate model or dataset for all applications. In this repository, we include software and the corresponding assessments of various global climate models (GCMs) from the Coupled Model Intercomparison Project (CMIP6), evaluating their skills with respect to the historical climate and comparing of their future projections of climate change. We focus on variables that directly affect the energy system including the representation of extreme values that can drive grid resilience events. The objective of this repository is not to recommend the best climate model and dataset for a given analysis, but instead to provide a reference to facilitate the selection of climate models and datasets in subsequent work.

Buster, Grant↗

Global Climate

The 2019 global lower tropospheric temperature (LTT), which encompasses the atmosphere from the surface to ~10 km, ranked second warmest in seven datasets and first or third in the remaining two (Fig. 2.7). These records extend back to 1958 using radiosonde (balloon-borne instrumentation) data and one reanalysis dataset (JRA55), which demonstrate reasonable agreement with the 40+ year satellite record (since late 1978) and two other reanalysis datasets (since 1979 and 1980, ERA5 and MERRA2, respectively). Aweak El Niño contributed to increased global temperatures as 2019 values were +0.44° to +0.68°C higher than the 1981–2010 average (depending on the dataset), being just slightly cooler (~0.07°C on average) than the record warm year of 2016. At least fourof the five globally complete datasets (ERA5, MERRA2, JRA55, RSS, UAH) recorded each of the four months—June, September, November, and December—as experiencing their warmest monthly global LTT.

58 GEOSCIENCES↗

Characterizing Sierra Nevada Snowpack Using Variable-Resolution CESM

The location, timing, and intermittency of precipitation in California make the state integrally reliant on winter-season snowpack accumulation to maintain its economic and agricultural livelihood. Of particular concern is that winter-season snowpack has shown a net decline across the western United States over the past 50 years, resulting in major uncertainty in water-resource management heading into the next century. Cutting-edge tools are available to help navigate and preemptively plan for these uncertainties. This paper uses a next-generation modeling technique-variable-resolution global climate modeling within the Community Earth System Model (VR-CESM)-at horizontal resolutions of 0.125° (14 km) and 0.25° (28 km). VR-CESM provides the means to include dynamically large-scale atmosphere-ocean drivers, to limit model bias, and to provide more accurate representations of regional topography while doing so in a more computationally efficient manner than can be achieved with conventional general circulation models. This paper validates VR-CESM at climatological and seasonal time scales for Sierra Nevada snowpack metrics by comparing them with the "Daymet," "Cal-Adapt," NARR, NCEP, and North American Land Data Assimilation System (NLDAS) reanalysis datasets, the MODIS remote sensing dataset, the SNOTEL observational dataset, a standard-practice global climate model (CESM), and a regional climate model (WRF Model) dataset. Overall, given California's complex terrain and intermittent precipitation and that both of the VR-CESM simulations were only constrained by prescribed sea surface temperatures and data on sea ice extent, a 0.68 centered Pearson product-moment correlation, a negative mean SWE bias of < 7 mm, an interquartile range well within the values exhibited in the reanalysis datasets, and a mean December-February extent of snow cover that is within 7% of the expected MODIS value together make apparent the efficacy of the VR-CESM framework.

54 ENVIRONMENTAL SCIENCES↗

An Intercomparison of GCM and RCM Dynamical Downscaling for Characterizing the Hydroclimatology of California and Nevada

Abstract Dynamical downscaling is a widely used technique to properly capture regional surface heterogeneities that shape the local hydroclimatology. However, in the context of dynamical downscaling, the impacts on simulation fidelity have not been comprehensively evaluated across many user-specified factors, including the refinements of model horizontal resolution, large-scale forcing datasets, and dynamical cores. Two global-to-regional downscaling methods are used to assess these: specifically, the variable-resolution Community Earth System Model (VR-CESM) and the Weather Research and Forecasting (WRF) Model with horizontal resolutions of 28, 14, and 7 km. The modeling strategies are assessed by comparing the VR-CESM and WRF simulations with consistent physical parameterizations and grid domains. Two groups of WRF Models are driven by either the NCEP reanalysis dataset (WRF_NCEP) or VR-CESM7 results (WRF_VRCESM) to evaluate the effects of large-scale forcing datasets. The simulated hydroclimatologies are compared with reference datasets for key properties including total precipitation, snow cover, snow water equivalent (SWE), and surface temperature. The large-scale forcing datasets are critical to the WRF simulations of total precipitation but not surface temperature, controlled by the wind field and atmospheric moisture transport at the ocean boundary. No significant benefit is found in the regional average simulated hydroclimatology by increasing horizontal resolution refinement from 28 to 7 km, probably due to the systematic biases from the diagnostic treatment of rainfall and snowfall in the microphysics scheme. The choice of dynamical core has little impact on total precipitation but significantly determines simulated surface temperature, which is affected by the snow-albedo feedback in winter and soil moisture estimations in summer.

54 ENVIRONMENTAL SCIENCES↗