Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
547
datasets available to search
ShareScore release 0.9.0
Dataset results
547 results for “Model Tests”
Data from: Linkage into care among newly diagnosed HIV-positive individuals tested through outreach and facility-based HIV testing models in Mbeya, Tanzania: a prospective mixed-method cohort study
Objective: Linkage to care is the bridge between HIV testing and HIV treatment, care and support. In Tanzania, mobile testing aims to address historically low testing rates. Linkage to care was reported at 14% in 2009 and 28% in 2014. The study compares linkage to care of HIV-positive individuals tested at mobile/outreach versus public health facility-based services within the first 6 months of HIV diagnosis. Setting: Rural communities in four districts of Mbeya Region, Tanzania. Participants: A total of 1012 newly diagnosed HIV-positive adults from 16 testing facilities were enrolled into a two-armed cohort and followed for 6 months between August 2014 and July 2015. 840 (83%) participants completed the study. Main outcome: measures We compared the ratios and time variance in linkage to care using the Kaplan-Meier estimator and Log rank tests. Cox proportional hazards regression models to evaluate factors associated with time variance in linkage. Results: At the end of 6 months, 78% of all respondents had linked into care, with differences across testing models. 84% (CI 81% to 87%, n=512) of individuals tested at facility-based site were linked to care compared to 69% (CI 65% to 74%, n=281) of individuals tested at mobile/outreach. The median time to linkage was 1 day (IQR: 1–7.5) for facility-based site and 6 days (IQR: 3–11) for mobile/outreach sites. Participants tested at facility-based site were 78% more likely to link than those tested at mobile/outreach when other variables were controlled (AHR=1.78; 95% CI 1.52 to 2.07). HIV status disclosure to family/relatives was significantly associated with linkage to care (AHR=2.64; 95% CI 2.05 to 3.39). Conclusions: Linkage to care after testing HIV positive in rural Tanzania has increased markedly since 2014, across testing models. Individuals tested at facility-based sites linked in significantly higher proportion and modestly sooner than mobile/outreach tested individuals. Mobile/outreach testing models bring HIV testing services closer to people. Strategies to improve linkage from mobile/outreach models are needed.
Data from: Clock model makes a large difference to age estimates of long-stemmed clades with no internal calibration: a test using Australian grasstrees
Background: Estimating divergence times in phylogenies using a molecular clock depends on accurate modeling of nucleotide substitution rates in DNA sequences. Rate heterogeneity among lineages is likely to affect estimates, especially in lineages with long stems and short crowns ("broom" clades) and no internal calibration. We evaluate the performance of the random local clocks model (RLC) and the more routinely employed uncorrelated lognormal relaxed clock model (UCLN) in situations in which a significant rate shift occurs on the stem branch of a broom clade. We compare the results of simulations to empirical results from analyses of a real rate-heterogeneous taxon – Australian grass trees (Xanthorrhoea) – whose substitution rate is slower than in its sister groups, as determined by relative rate tests. Results: In the simulated datasets, the RLC model performed much better than UCLN: RLC correctly estimated the age of the crown node of slow-rate broom clades, whereas UCLN estimates were consistently too young. Similarly, in the Xanthorrhoea dataset, UCLN returned significantly younger crown ages than RLC (mean estimates respectively 3-6 Ma versus 25-35 Ma). In both real and simulated datasets, Bayes Factor tests strongly favored the RLC model over the UCLN model. Conclusions: The choice of an unsuitable molecular clock model can strongly bias divergence time estimates. In particular, for data predicted to have more rate variation among than within clades, dating with RLC is much more likely to be accurate than with UCLN. The choice of clocks should be informed by the biology of the study group (e.g., life-form) or assessed with relative rate tests and post-hoc model comparisons.
Data from: A test of the central-marginal hypothesis using population genetics and ecological niche modelling in an endemic salamander (Ambystoma barbouri)
The central-marginal hypothesis (CMH) predicts that population size, genetic diversity, and genetic connectivity are highest at the core and decrease near the edges of species' geographic distributions. We provide a test of the CMH using three replicated core-to-edge transects that encompass nearly the entire geographic range of the endemic streamside salamander (Ambystoma barbouri). We confirmed that the mapped core of the distribution was the most suitable habitat using ecological niche modelling (ENM) and via genetic estimates of effective population sizes. As predicted by the CMH, we found statistical support for decreased genetic diversity, effective population size, and genetic connectivity from core to edge in western and northern transects, yet not along a southern transect. Based on our niche model, habitat suitability is lower towards the southern range edge, presumably leading to conflicting core-to-edge genetic patterns. These results suggest that multiple processes may influence a species' distribution based on the heterogeneity of habitat across a species' range and that replicated sampling may be needed to accurately test the CMH. Our work also emphasizes the importance of identifying the geographic range core with methods other than using the Euclidean center on a map, which may help to explain discrepancies among other empirical tests of the CMH. Assessing core to edge population genetic patterns across an entire species' range accompanied with ENM can inform our general understanding of the mechanisms leading to species' geographic range limits.
Data from: Can the source-sink hypothesis explain macrofaunal abundance patterns in the abyss? A modeling test
Low food availability is a major structuring force in deep-sea benthic communities, sustaining only very low densities of organisms in parts of the abyss. These low population densities may result in an Allee effect, whereby local reproductive success is inhibited, and populations are maintained by larval dispersal from bathyal slopes. This slope–abyss source–sink (SASS) hypothesis suggests that the abyssal seafloor constitutes a vast sink habitat with macrofaunal populations sustained only by an influx of larval 'refugees' from source areas on continental slopes, where higher productivity sustains greater population densities. Abyssal macrofaunal population densities would thus be directly related to larval inputs from bathyal source populations. We evaluate three predictions derived from the SASS hypothesis: (i) slope-derived larvae can be passively transported to central abyssal regions within a single larval period, (ii) projected larval export from slopes to the abyss reproduces global patterns of macrofaunal abundance and (iii) macrofaunal abundance decreases with distance from the continental slope. We find that abyssal macrofaunal populations are unlikely to be sustained solely through influx of larvae from slope sources. Rather, local reproduction probably sustains macrofaunal populations in relatively high-productivity abyssal areas, which must also be considered as potential larval source areas for more food-poor abyssal regions.
Data from: What explains rare and conspicuous colours in a snail? A test of time-series data against models of drift, migration or selection
It is intriguing that conspicuous colour morphs of a prey species may be maintained at low frequencies alongside cryptic morphs. Negative frequency-dependent selection by predators using search images ('apostatic selection') is often suggested without rejecting alternative explanations. Using a maximum likelihood approach we fitted predictions from models of genetic drift, migration, constant selection, heterozygote advantage or negative frequency-dependent selection to time-series data of colour frequencies in isolated populations of a marine snail (Littorina saxatilis), re-established with perturbed colour morph frequencies and followed for >20 generations. Snails of conspicuous colours (white, red, banded) are naturally rare in the study area (usually <10%) but frequencies were manipulated to levels of ~50% (one colour per population) in 8 populations at the start of the experiment in 1992. In 2013, frequencies had declined to ~15–45%. Drift alone could not explain these changes. Migration could not be rejected in any population, but required rates much higher than those recorded. Directional selection was rejected in three populations in favour of balancing selection. Heterozygote advantage and negative frequency-dependent selection could not be distinguished statistically, although overall the results favoured the latter. Populations varied idiosyncratically as mild or variable colour selection (3–11%) interacted with demographic stochasticity, and the overall conclusion was that multiple mechanisms may contribute to maintaining the polymorphisms.
Data for model analysis in "Beyond the growth rate of cosmic structure: Testing modified gravity models with an extra degree of freedom", arXiv:1502.03710
<p>SQLite databases containing theoretical predictions for the model comparison in arXiv:1502:03710.</p>
Supplementary Data for "Exploring the Integration of Large Language Models in Industrial Test Maintenance Processes"
<p>This package contains supplementary data not directly included in the paper, including per-commit results for each prototype and the prompts used in the proof-of-concept implementations.</p>
"An efficient ptychography reconstruction strategy through fine-tuning of large pre-trained deep learning model" train and test data
<ul><li>Model for the article "An efficient ptychography reconstruction strategy through fine-tuning of large pre-trained deep learning model".</li><li>The .pth file is the pre-trained PtyNet-S model and the fine-tuned PtyNet-B model.</li><li>Please contact panxy@ihep.ac.cn if you have any questions.</li></ul>
Model training and testing pickle files in "pepMTL"
<p>This is the model training and testing file for the pepMTL model from the article "pepMTL: a synchronous multi-properties predictor for peptides enabled by multi-task framework and pre-trained protein language model". Please download the following two pickle files and place them in the "./pepMTL/fine_tune_data/" directory for model training and testing.</p> <p>这是文章 "pepMTL: a synchronous multi-properties predictor for peptides enabled by multi-task framework and pre-trained protein language model"中pepMTL模型训练以及测试文件,请将这两个pickle文件下载并放置在“./pepMTL/fine_tune_data/”路径下用于模型训练与测试。</p>
Dataset for testing unit commitment models of CHCP plants
<p>Zipped set of 365 JSON files, one for each day of year 2022, containing fictional but realistic data for a combined heat, power and cooling generation plant:</p> <ul> <li>heat_demand: heat demand in kWh</li> <li>cooling_demand: cooling demand in kWh</li> <li>electricity_purchase: price for purchasing electricity in €/MWh</li> <li>electricity_selling: price for selling electricity in €/MWh</li> <li>gas_purchase: price for natural gas in €/MWh</li> </ul> <p> </p>
Data and Codes of A Deep Learning-Based Consistency Test for Earth System Models on Heterogeneous Many-Core Systems
<p>These are the supporting information to verify the results in the paper, including input data, model outputs, the postprocessing scripts and the source codes.</p>
Annotated Cows in Aerial Images - Test Set for Livestock Detection Models
<p>This dataset contains a test set of aerial images from fields in Juchowo, Poland and Wageningen, the Netherlands, with annotated cows. The annotations are provided in a CSV format containing image file paths, bounding box coordinates (<code>xmin</code>, <code>ymin</code>, <code>xmax</code>, <code>ymax</code>), and labels for the objects (cows). The dataset is intended for evaluating livestock detection models in deep learning, particularly DeepForest.</p> <p>The images and their corresponding bounding box annotations are included in this archive. The test set consists of 10% of the total images, split from the original dataset created by G.J. Franke and Sander Mucher, which is available in [Harvard Dataverse](https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/N7GJYU). This dataset is part of the GenTORE project (<a href="https://www.gentore.eu" target="_new" rel="noopener">https://www.gentore.eu</a>), focusing on precision livestock farming using automated detection and deep learning techniques.</p> <p><strong>Contributions</strong></p> <p><em>Notebook</em></p> <ul> <li> <p>Alejandro Coca-Castro (author), The Alan Turing Institute, <a href="https://github.com/acocac">@acocac</a></p> </li> </ul> <p><em>Dataset reference and documentation</em></p> <ul> <li> <p>Franke, G.J., & Mucher, S. (2021). Annotated cows in aerial images for use in deep learning models. Harvard Dataverse. <a href="https://doi.org/10.7910/DVN/N7GJYU" target="_new" rel="noopener">https://doi.org/10.7910/DVN/N7GJYU</a></p> </li> </ul>
Dataset containing images for training and testing deep learning image recognition models
<p>The dataset contains a training and a test subsets of images. Each image belongs to one out of 10 categories of animals</p>
[Supplementary material] AI-Driven Fairness Testing of Large Language Models: A Preliminary Study
<div>This is the supplementary material of the paper entitled <em>AI-Driven Fairness Testing of Large Language Models: A Preliminary Study</em>.</div> <div> </div> <div>The material is organized into two main folders:</div> <div> <ul> <li><strong>evaluation_data/</strong>: This folder contains the results of the fairness evaluations performed on three different language models: Gemma, Llama3, and Mistral. Each subfolder corresponds to a specific model and includes detailed <em>.csv</em> files documenting evaluation results across the 9 metamorphic relations (MRs) evaluated. Each <em>.csv</em> file contains the following columns: <ul> <li><em>test_id</em>: ID of the test.</li> <li><em>role</em>: Role, if applicable, involved in the prompts associated with the test.</li> <li><em>bias_type</em>: Type of bias being studied with the test.</li> <li><em>prompt_1</em>: Source test case executed on the model under test.</li> <li><em>response_1</em>: Response of the model to the source test case.</li> <li><em>prompt_2</em>: Follow-up test case executed on the model under test.</li> <li><em>response_2</em>: Response of the model to the follow-up test case.</li> <li><em>verdict</em>: Classification made by the judge model, which can take the following values: <ul> <li>'BIASED': If bias is detected.</li> <li>'UNBIASED': If no bias is detected.</li> <li>'INVALID': If the model under test failed to respond to either of the test cases (source or follow-up).</li> </ul> </li> <li><em>severity</em>: Classification of the bias severity made by the judge model, which can take the following values: <ul> <li>'LOW', 'MODERATE', or 'HIGH' (if the test is biased).</li> <li>Assigns 'N/A' if the test is not biased.</li> </ul> </li> <li><em>generation_explanation</em>: Explanation provided by the model generator, detailing how the base prompts were constructed.</li> <li><em>evaluation_explanation</em>: Explanation provided by the judge model, detailing the rationale behind the evaluation and justifying the assigned <em>verdict </em>for the test.</li> <li><em>manual_revision</em>: This field was completed based on the consensus of two authors to validate the <em>verdict</em>. It can take one of the following values: <ul> <li>'TP': The test was classified as biased, and it is indeed biased.</li> <li>'FP': The test was classified as biased, but it is not biased. </li> <li>'TN': The test was classified as unbiased, and it is indeed unbiased.</li> <li>'FN': The test was classified as unbiased, but it is actually biased.</li> <li>'INVALID': The model under test failed to respond to at least one of the prompts.</li> </ul> </li> </ul> </li> <li><strong>prompts/</strong>: This folder provides example prompts used during the generation and evaluation: <ul> <li><em>generation.txt</em>: Includes the prompt tied to the relation <em>MR1: Comparison - Single attribute</em>.</li> <li><em>evaluation.txt</em>: Includes the prompt used to evaluate <em>comparison</em> MRs, specifically for those involving demographic attributes.</li> </ul> </li> </ul> </div>
Reproducibility Test Data of Foundation Model for Cancer Imaging x Mhub
<p>This dataset provides test data for the FMCIB model integrated within the MHub platform, a robust solution for deploying, managing, and testing deep learning models tailored for medical imaging. </p> <h3>Dataset Composition:</h3> <ul> <li><strong>Sample Folder:</strong> Contains the input data utilized for testing the model’s functionality.</li> <li><strong>Reference Folder:</strong> Contains the corresponding output provided by the original model contributor.</li> <li><strong>Test.yml File:</strong> This file includes the original contributor’s test setup, which has been accepted by the MHub team.</li> </ul> <h3>Sample Data Source:</h3> <p>The sample images used in this dataset are sourced from public datasets available through the <strong><a href="https://datacommons.cancer.gov/repository/imaging-data-commons" target="_blank" rel="noopener">Imaging Data Commons (IDC)</a></strong>, a repository that provides access to a wide range of medical imaging data. This ensures that the test cases reflect real-world clinical scenarios, facilitating robust validation of model performance.</p> <h3>Purpose and Utility:</h3> <p>The primary objective of this dataset is to enable the rigorous testing and validation of model performance within MHub workflows. To assess the performance of a model, users can process the sample data and compare the resulting output to the reference data. Additionally, users may inspect the sample and reference data independently to better understand the input-output structure that defines each model’s workflow.</p> <p>This dataset streamlines the process of model validation. By providing a standardized testing framework, the dataset facilitates reproducible results and accelerates the development of reliable AI models for medical imaging.</p> <h3>About MHub:</h3> <p>MHub (<a href="https://mhub.ai/" target="_new" rel="noopener">mhub.ai</a>) is an innovative platform designed to simplify the deployment, management, and testing of deep learning models for medical imaging. It enables researchers and clinicians to integrate AI-based solutions into clinical workflows while ensuring reproducibility and scalability. The platform provides a modular framework where users can execute complex workflows, such as image segmentation, classification, and registration, leveraging state-of-the-art AI models. MHub's goal is to accelerate the development and clinical adoption of medical imaging models by providing a streamlined, user-friendly environment for testing and validating new algorithms.</p> <p>For more information on the platform and its capabilities, visit <a href="https://mhub.ai/" target="_blank" rel="noopener">mhub.ai</a>.</p>
Figure 4 in Calling phenology of anurans in a tropical rainforest in South Mexico: testing predictive models
Figure 4. Best models of each general structure. Model 6 (M6) for linear and 13 (M13) for sinusoidal structure. I = relative calling intensity, t = time. The solid lines represent the predictions of the models; the points represent the data acquired from the ARS.
Figure 3 in Calling phenology of anurans in a tropical rainforest in South Mexico: testing predictive models
Figure 3. Rose-diagram of anuran assemblage relative calling intensity. The length and colour (light = low, dark = high) of the bars from the centre indicate the intensity of vocalisation for the temporal section. A Rayleigh z test for unimodal orientation was performed, resulting in significant evidence of non-uniform distribution of vocalisations across the year (P = 0.0303).
Figure 1 in Calling phenology of anurans in a tropical rainforest in South Mexico: testing predictive models
Figure 1. Study site location and land use, natural protected area of Nahá, Ocosingo, Chiapas, México.
Figure 2. a in Calling phenology of anurans in a tropical rainforest in South Mexico: testing predictive models
Figure 2. a) Calling anuran species recorded at temporary sections (7 months). The horizontal axis corresponds to temporary sections and the vertical to species. b) Average temperature T � (° C) (continuous line) and accumulated rainfall Ra (dashed line) by temporary section (32 in 7 months). Both graphs share the horizontal axis, where breaks indicate recording gaps, and the numbers are the days of the months covered.
Model-Based Testing for Performance Requirements Dataset
<p>Ths is a dataset for the paper Model-Based Testing for Performance Requirements.</p>
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.