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.
1,663
datasets available to search
ShareScore release 0.9.0
Dataset results
1,663 results for “bias”
EVIDENT H2020 - Average Price Bias Dataset
<p>The average price bias choice quasi-experiment is designed to elicit consumers’ perceptions about different pricing schemes. The experiment aims to correlate the findings with participants’ characteristics, potential behavioural biases and the participants’ financial and environmental literacy levels.</p> <p>The experiment consists of five discrete key sections: 1) a section about participant’s demographic data, 2) a small set of questions related to behavioural biases, 3) a set of questions related to financial literacy, 4) a section with questions related to environmental literacy and 5) the choice experiment about price perceptions.</p> <p>Section 5 presents a hypothetical scenario about the participant’s yearly energy consumption and several pricing tariff options. The participant has to choose a pricing tariff they think is the most cost-effective. There are six (1-6) broader cases, each including four subcategories (a-d). The further the case is from the beginning, the more complicated it is.</p> <p>The implementation of the experiment is as follows:</p> <p>Step 1. The participant first receives the following message: “Assuming that your yearly energy consumption is exactly 6,000 kWh, which one of the following tariffs would you choose as the most cost-effective?”</p> <p>Step 2. Each participant will be asked to participate in only 2 cases (all subcategories of each case are included). A case will be randomly chosen from cases 1-3 (simple case) and a second random choice will be made from cases 4-6 (complex case). Thus, all participants will answer a simple and a complicated set of questions.</p> <p>Step 3. The participant receives the first set of choices.</p> <p>If the participant answers correctly, he receives the next subcategory's choice set. If he answers false, he gets the next set of choices within the same subcategory. Thus, as soon the participant answers correctly, he skips the following set of choices and moves to the next subcategory. For a participant answering correctly, this will be a short survey. However, for someone answering wrong, the survey will last longer.</p> <p>More information can be found on the public deliverables of the EVIDENT project <a href="https://evident-h2020.eu/deliverables/">https://evident-h2020.eu/deliverables/</a>. More specifically, the experiment's theoretical framework and motivation are described in deliverable <strong>D1.2</strong> <a href="https://evident-h2020.eu/wp-content/uploads/2021/12/EVIDENT_D1.2_Assessing_behavioural_biases_and_financial_literacy.pdf">Assessing behavioural biases and financial literacy</a>, in section 5 while the final design is reported in <strong>D3.2</strong> <a href="http://evident-h2020.eu/wp-content/uploads/2023/01/EVIDENT_D3.2_Implementation-of-preparatory-actions-for-RCT-surveys-and-serious-game.pdf">Implementation of preparatory actions for RCT, surveys and serious game</a>.</p>
A Simple Strategy to Eliminate Glycation Bias in the Relative Quantification of Protein N-glycosylation
<p><strong>Contents</strong></p> <p>Supplementary files for “A Simple Strategy to Eliminate Glycation Bias in the Relative Quantification of Protein <em>N</em>-glycosylation” by Esser-Skala et al (2020):</p> <ul> <li> <p><em>cafog_raw_data.tar.gz</em>: Raw data for denosumab and NISTmAb.</p> </li> <li> <p><em>cafog_source_code.zip</em>: Supplementary file 1 from the manuscript. This ZIP archive contains the source code of CAFOG.</p> </li> <li> <p><em>data.zip</em>: Supplementary file 2 from the manuscript. Files in this ZIP archive allow to reproduce all results presented in the manuscript.</p> </li> </ul> <p> </p> <p><strong>Changelog</strong></p> <ul> <li>1.1.0 – 2023-05-25 <ul> <li>added supplementary files 1 and 2 mentioned in the manuscript, since those were not published along with the manuscript</li> </ul> </li> <li>1.0.0 – 2020-01-30 <ul> <li>initial release</li> </ul> </li> </ul>
Hardware-biased quantum random numbers
<p>Bitstring from a hardware-biased quantum random number generator. The data is used in the numerical experiments for the publication "On the effects of biased quantum random numbers on the initialization of artificial neural networks" (<a href="http://arxiv.org/abs/2108.13329">arXiv:2108.13329</a>).</p> <p> </p>
Code and data for: Is habitat selection in the wild shaped by individual-level cognitive biases in orientation strategy?
<p>This repository is a companion to the manuscript "<em>Is habitat selection in the wild shaped by individual-level cognitive biases in orientation strategy?</em>" and is linked to <a href="https://github.com/CBeardsworth/Pheasant_OrientStrat_Habitat">Github</a>.</p> <p>For any questions about the code please contact Christine at <a href="mailto:c.e.beardsworth@gmail.com">c.e.beardsworth@gmail.com</a></p> <p>To use any data contained in this repository contact Joah at <a href="mailto:j.r.madden@exeter.ac.uk">j.r.madden@exeter.ac.uk</a> for permission.</p> <p>In this repository, we have included a run-through of the R analysis <a href="https://cbeardsworth.github.io/Pheasant_OrientStrat_Habitat/">here</a> to show the outputs of the analysis without the need to run the code. For those that might want to run the code themselves, we have included three R scripts (<a href="https://github.com/CBeardsworth/NavigationHabitat/blob/master/R">/R</a>) and their accompanying datasets (<a href="https://github.com/CBeardsworth/NavigationHabitat/blob/master/Data">/Data</a>). A description of the code and the data needed to run them is below:</p> <p><em>Cognition analysis and figs.R</em> = Run the cognition analysis for the first section of the manuscript and create the figures. For this, the datasets mazeData.csv (the learning trials) and mazeRotationResults.csv (the probe trial) are required. </p> <p><em>iSSA analysis and bootstrapping.R</em> = Run iSSA models and bootstrapping. This produces the datasets required for the next stage of analysis. For this code, the datasets habitat.grd (habitat information), atlas2018-strategy.csv (atlas data + id and strategy data for each bird) and FeederCoords2017_27700.csv (coordinates of feeder locations from 2017-2018) are required. The produced datasets are included in <a href="https://github.com/CBeardsworth/NavigationHabitat/blob/master/Data">/Data</a> therefore to run subsequent analyses, this code does not need to be run. To develop this code we relied heavily on the code included in the supplementary material of <a href="https://doi.org/10.1002/ece3.4823">Signer et al. (2019)</a> as well as an <a href="https://bsmity13.github.io/log_rss">online tutorial</a> from Brian J. Smith for calculating log-RSS.</p> <p><em>Habitat analysis and Figs.R</em> = Run the statistical models for the final section of the manuscript and create the figures. For this code, the datasets produced in the previous R script are required (habitatOrientation_coefs.csv and habitatOrientation_avail.csv). We have included <a href="https://github.com/CBeardsworth/NavigationHabitat/blob/master/Data">these datasets</a> so users do not need to run the iSSA analysis and bootstrapping.R script themselves. </p>
Experimental Data on Pragmatic, Constructive and Reconstructive Memory Influences on the Hindsight Bias
<p>After knowing how events turned out, we are quick to say ‘we knew it all along.’ Decades of research on hindsight bias have shown that outcome information biases what we later present as our original judgments. This experiment combined established between- and within-participant designs in a longitudinal study.</p>
Data set of the article: Language Bias in the Google Scholar Ranking Algorithm
<p>Data of investigation published in the article Cristòfol Rovira; Lluís Codina; Carlos Lopezosa Language Bias in the Google Scholar Ranking Algorithm. Future Internet, 2021, 13.</p> <p><strong>Abstract: </strong>The visibility of academic articles or conference papers depends on their being easily found in academic search engines, above all in Google Scholar. To enhance this visibility, search engine optimization (SEO) has been applied in recent years to academic search engines in order to optimize documents and, thereby, ensure they are better ranked in search pages (i.e., academic search engine optimization or ASEO). To achieve this degree of optimization, we first need to further our understanding of Google Scholar’s relevance ranking algorithm, so that, based on this knowledge, we can highlight or improve those characteristics that academic documents already present and which are taken into account by the algorithm. This study seeks to advance our knowledge in this line of research by determining whether the language in which a document is published is a positioning factor in the Google Scholar relevance ranking algorithm. Here, we employ a reverse engineering research methodology based on a statistical analysis that uses Spearman’s correlation coefficient. The results obtained point to a bias in multilingual searches conducted in Google Scholar with documents published in languages other than in English being systematically relegated to positions that make them virtually invisible. This finding has important repercussions, both for conducting searches and for optimizing positioning in Google Scholar, being especially critical for articles on subjects that are expressed in the same way in English and other languages, the case, for example, of trademarks, chemical compounds, industrial products, acronyms, drugs, diseases, etc.</p>
The Choice is Yours? How Algorithm Bias Impacts Fairness and Accessibility of Knowledge
<p><strong>Episode Summary</strong></p> <p>In this episode we talked about 'almighty' algorithms with Carlos Castillo, Lorenzo Porcaro, Marzieh Karimihaghighi, David Solans, and Francesco Fabbri from the Web Science & Social Computing Research Group, and the department of Engineering in Information & Communication Technologies, in Universitat Pompeu Fabra in Barcelona. We discussed how bias can enter into algorithm systems, how bias is measured, and what systems are impacted by it. </p> <p><strong>Episode Links</strong></p> <p><a href="https://www.upf.edu/web/wssc/">Web Science and Social Computing Research Group</a></p> <ul> <li><a href="https://www.upf.edu/web/etic/entry/-/-/24095/adscripcion/carlos-alberto-alejandro-castillo">Carlos Castillo</a></li> <li><a href="https://www.linkedin.com/in/marzieh-karimihaghighi-706b5554/?originalSubdomain=ir">Marzieh Karimihaghighi</a></li> <li><a href="https://www.linkedin.com/in/david-solans-noguero-48269b85/?originalSubdomain=es">David Solans</a></li> <li><a href="https://www.linkedin.com/in/francesco-fabbri/?originalSubdomain=it">Francesco Fabbri</a></li> <li><a href="https://www.linkedin.com/in/lorenzo-porcaro-7a8792b1/?originalSubdomain=es">Lorenzo Porcaro</a></li> </ul>
Data from: An analysis of mating biases in trees
<p><span><span><span><span><span><span><span><span><span><span><span>Assortative mating is a deviation from random mating based on phenotypic similarity. As it is much better studied in animals than in plants, we investigate for trees whether kinship of realized mating pairs deviates from what is expected from the set of potential mates and use this information to infer mating biases that may result from kin recognition and/or assortative mating. Our analysis covers twenty species of trees for which microsatellite data is available for adult populations (potential mates) as well as seed arrays. We test whether mean relatedness of observed mating pairs deviates from null expectations that only take pollen dispersal distances into account (estimated from the same dataset). This allows to identify elevated as well as reduced kinship among realized mating pairs, indicative of positive and negative assortative mating, respectively. The test is also able to distinguish elevated biparental inbreeding that occurs solely as a result of related pairs growing closer to each other from further assortativeness.</span></span></span></span></span></span></span></span></span></span></span></p> <p><span><span><span><span><span><span><span><span><span><span><span>Assortative mating in trees appears potentially common but not ubiquitous: nine data sets show mating bias with elevated inbreeding, nine do not deviate significantly from the null expectation, and two show mating bias with reduced inbreeding.</span></span></span></span></span></span></span></span></span></span></span></p> <p><span><span><span><span><span><span><span><span><span><span><span>While our datasets lack direct information on phenology, our investigation of the phenological literature for each species identifies flowering phenology as a potential driver of positive assortative mating (leading to elevated inbreeding) in trees. Since active kin recognition provides an alternative hypothesis for these patterns, we encourage further investigations on the processes and traits that influence mating patterns in trees.</span></span></span></span></span></span></span></span></span></span></span></p>
Test Collection Reliability: A Study of Bias and Robustness to Statistical Assumptions via Stochastic Simulation
<p>This archive contains the simulated collections, their diagnosis data, and the estimates of accuracy. For the full code and description, please refer to https://github.com/julian-urbano/irj2015-reliability</p>
abarbour/pbe-data: PBO BSM data, peak values, and regression biases -- updated with revised data and CRUST1.0 terms
<p>Two new tables are included here, each containing site and source classification terms from the CRUST1.0 global velocity model (see http://igppweb.ucsd.edu/~gabi/crust1.html)</p> <p>Dataset from Barbour and Crowell (2017), Dynamic Strains for Earthquake Source Characterization, <em>Seismol. Res. Lett.</em>, <strong>88</strong> (2A) 354-370; DOI: 10.1785/0220160155</p>
Attentional Bias for Uncertain Cues of Shock in Human Fear Conditioning: Evidence for Attentional Learning Theory
<p>Eye tracking data and statistical analysis of:</p> <p>Koenig, S., Uengoer, M., & Lachnit, H. (2017). Attentional bias for uncertain cues of shock in human fear conditioning: Evidence for attentional learning theory. Frontiers in Human Neuroscience. doi: 10.3389/fnhum.2017.00266.</p> <p>Abstract: We conducted a human fear conditioning experiment in which three different color cues were followed by an aversive electric shock on 0, 50, and 100% of the trials, and thus induced low (L), partial (P), and high (H) shock expectancy respectively. The cues differed with respect to the strength of their shock association (L < P < H) and the uncertainty of their prediction (L < P > H). During conditioning we measured pupil dilation and ocular fixations to index differences in the attentional processing of the cues.<br> After conditioning, the shock-associated colors were introduced as irrelevant distracters during visual search for a shape target while shocks were no longer administered and we analyzed the cues’ potential to capture and hold overt attention automatically.<br> Our findings suggest that fear conditioning creates an automatic attention bias for the conditioned cues that depends on their correlation with the aversive outcome. This bias was exclusively linked to the strength of the cues’ shock association for the early<br> attentional processing of cues in the visual periphery, but additionally was influenced by the uncertainty of the shock prediction after participants fixated on the cues. These findings are in accord with attentional learning theories that formalize how associative learning shapes automatic attention.</p>
Data from "Asymmetries in behavioral and neural responses to spectral cues demonstrate the generality of auditory looming bias"
<p>Supporting material for Baumgartner et al. (2017): "Asymmetries in behavioral and neural responses to spectral cues demonstrate the generality of auditory looming bias" in Proc Natl Acad Sci USA; www.pnas.org/cgi/doi/10.1073/pnas.1703247114</p>
Bias Correction of CRCM5-LE for Hydrological Bavaria
<p>The frequency and intensity of extreme hydrometeorological events are anticipated to rise as a result of climate change. For precise analysis, especially in low-flow assessments, it is crucial to have data on precipitation and temperature with high spatial and sub-daily resolution. However, such data is often lacking in both density and duration. The <a href="https://www.climex-project.org/">ClimEx-II </a>project (Climate Change and Hydrological Extreme Events 2nd Phase) is dedicated to enhancing our understanding of these shifts in hydrological extremes.</p> <p>The Canadian Regional Climate Model version 5 Large Ensemble (CRCM5-LE; Leduc et al. (2019)) under RCP 8.5 builds the climatic boundary conditions for the hydrological modelling. The ensemble covers a European and a North American domain, each comprising 50 members from 1951 to 2100. The SDCLIREF v2 (Lehr- und Forschungseinheit für physische Geographie und komplexe Umweltsysteme 2024), a sub-daily (3h), high-resolution (500m) data set for the domain of Bavaria and hydrologically important neighbouring catchments marks the reference data set. </p> <p>In ClimEx-II, bias correction is a crucial step before downscaling (regional) climate model simulations to higher resolutions as it adjusts local inconsistencies in the climate model. The corrected and downscaled meteorological inputs can be used to drive a hydrological model (Emami and Koch 2018, Fang et al. 2015). The quality of the corrected data depends on the method used. Therefore, this data set comprises a comparison of the input and output data from three different bias correction methods, UBC (Cannon et al., 2015), MBCn (Cannon 2018) and VBC (Funk et al., 2024) for three diverse climate regions in Bavaria:</p> <ul> <li>Fränkische Saale Salz is a franconian catchment</li> <li>Iller Kempten is a pre-alpine catchment</li> <li>Hart an der Ziller is an alpine catchment</li> </ul> <p>Each catchment comprises six to seven grid cells of a 12 km resolution. Five climate variables of hydrological importance are corrected in a 3-hourly temporal resolution per grid cell:</p> <ul> <li>Near-Surface Dewpoint Temperature in °C (<em>dew</em>)</li> <li>Precipitation in kg/m2 (<em>pr</em>)</li> <li>Surface Downwelling Shortwave Radiation in W/m2 (<em>rsds</em>)</li> <li>Near-Surface Wind Speed in m/s (<em>sfcWind</em>)</li> <li>Near-Surface Air Temperature in °C (<em>tas</em>)</li> </ul> <p>The environment in each file comprises the inputs to the bias correction</p> <ul> <li><strong>mp_dts</strong>: CRCM5-LE model data during the projection period (2011-2030) before correction</li> <li><strong>mc_dts</strong>: CRCM5-LE model data during calibration period (1991-2010)</li> <li><strong>oc_dts</strong>: SDCLIREF v2 reference data during calibration period (1991-2010)</li> </ul> <p>,the outputs from the bias correction comparison during the projection period</p> <ul> <li><strong>vbc</strong>: CRCM5-LE model data during the projection period (2011-2030) after correction by VBC</li> <li><strong>mbcn</strong>: CRCM5-LE model data during the projection period (2011-2030) after correction by MBCn</li> <li><strong>ubc</strong>: CRCM5-LE model data during the projection period (2011-2030) after correction by UBC</li> </ul> <p>and the held-out reference data for validation</p> <ul> <li><strong>op_dts</strong>: SDCLIREF v2 reference data during the projection period (2011-2030).</li> </ul> <p>Each of the above-presented data sets consists of six columns. The five climate variables are indexed by their abbreviations. The sixth column <em>time</em> contains a string marking the respective timestamp. All CRCM5-LE model data contain a seventh column indicating the respective ensemble member. The evaluation results by Wasserstein Distance and Model Correction Inconsistency from Funk et al. (2024) are captured in the three additional files starting with <em>06_*</em>.</p>
Observational Bias and Young Massive Cluster Characterisation II. Can Gaia accurately observe young clusters and associations?
<p>Field-of-View for synthetic Gaia observations of clusters Orion-type-3, Orion-type-5.5 and Wd2-type presented in Buckner et al. (2023).</p><p>Files contain both simulation and field stars along the Line-of-Sight (l = 270^o, b = 0^o) for the clusters when placed at 500pc, 2500pc and 4300pc distances.</p><p>The original simulation files are included for reference.</p><p>Included README files provide more detailed descriptions of the files.</p>
Determining non-significant bits on a C++ implementation of the LeNet-5 convolutional neural network to be used for storing error correcting codes to protect weights and biases. Robustness assessment of the network after integrating the proposed codes.
<p>The architecture of the LeNet-5 convolutional neural network (CNN) was defined by LeCun in its paper "Gradient-based learning applied to document recognition" (<a href="https://ieeexplore.ieee.org/document/726791">https://ieeexplore.ieee.org/document/726791</a>) to classify images of hand written digits (MNIST dataset).</p><p>This architecture has been customized to use Rectified Linear Unit (ReLU) as activation functions instead of Sigmoid.</p><p>It consists of the following layers:</p><ul><li><strong>conv1</strong>: Convolution 2D, 1 input channel (28x28), 3 output channels (28x28), kernel size 5, stride 1, padding 2.</li><li><strong>relu1</strong>: Rectified Linear Unit (3@28x28).</li><li><strong>max1</strong>: Subsampling buy max pooling (3@14x14).</li><li><strong>conv2</strong>: Convolution 2D, 3 input channels (14x14), 6 output channels (14x14), kernel size 5, stride 1, padding 2.</li><li><strong>relu2</strong>: Rectified Linear Unit (6@14x14).</li><li><strong>max2</strong>: Subsampling buy max pooling (6@7x7).</li><li><strong>fc1</strong>: Fully connected (294, 147)</li><li><strong>fc2</strong>: Fully connected (147, 10)</li></ul><p>The fault hypotheses for this work include the occurrence of:</p><ul><li><strong>S0</strong>/<strong>S1</strong>: multiple adjacent stuck-at-0 and stuck-at-1 faults to determine the least significant bits of weights and biases that could be used to store the proposed error correcting codes.</li><li><strong>BF</strong>: single, double, and triple bit-flip faults to assess the robustness of the considered CNN</li></ul><p>In the memory cells containing all the parameters of the CNN: </p><ul><li><strong>w</strong>: weights (float32)</li><li><strong>b</strong>: biases (float32)</li></ul><p>All the images (10000) from the MNIST dataset have been used as workload.</p><p>The weights and biases of the LeNet-5 architecture have been protected using six different error correcting codes that have been deployed in the least significant bits of these elements.</p><p>The parity check matrices (H = P I) that define these ECCs are:</p><ul><li><strong>SEC(32, 26)</strong> (Hamming) under a <i>classic policy </i>(see methodology below):</li></ul><p><i> 11010010001000011101101000 100000</i></p><p><i> 10101001000100011011010100 010000</i></p><p><i> 01100100100010010110110010 001000</i></p><p><i> 00011100010001001110001101 000100</i></p><p><i> 00000011110000100001111011 000010</i></p><p><i> 00000000001111100000000111 000001</i></p><ul><li><strong>SEC(23, 18)</strong> (Hamming) under a <i>conservative policy</i> (see methodology below):</li></ul><p><i> 111100001111000000 10000</i></p><p><i> 110011101000111000 01000</i></p><p><i> 101011010100100110 00100</i></p><p><i> 010110110010010101 00010</i></p><p><i> 001101110001001011 00001</i></p><ul><li><strong>SEC(13, 9)</strong> (Hamming) under an <i>aggressive policy </i>(see methodology below):</li></ul><p><i> 110111000 1000</i></p><p><i> 101100110 0100</i></p><p><i> 011010101 0010</i></p><p><i> 111001011 0001</i></p><ul><li><strong>DEC(32, 21)</strong> (low redundancy and reduced overhead DEC) under a <i>classic policy </i>(see methodology below):</li></ul><p><i> 111000011001010010000 10000000000</i></p><p><i> 110110000011101000000 01000000000</i></p><p><i> 101011000110000010001 00100000000</i></p><p><i> 100101101000110001000 00010000000</i></p><p><i> 011010101100100000100 00001000000</i></p><p><i> 010101010100001001010 00000100000</i></p><p><i> 001100110010010100100 00000010000</i></p><p><i> 000011110001000110010 00000001000</i></p><p><i> 000000001111001101001 00000000100</i></p><p><i> 000000000000111100111 00000000010</i></p><p><i> 000000000000000011111 00000000001</i></p><ul><li><strong>DEC(28, 18)</strong> (low redundancy and reduced overhead DEC) under a <i>conservative policy </i>(see methodology below):</li></ul><p><i> 111111000000000000 1000000000</i></p><p><i> 110100111100000000 0100000000</i></p><p><i> 110000100011110000 0010000000</i></p><p><i> 001110010011001100 0001000000</i></p><p><i> 101100001010101010 0000100000</i></p><p><i> 010001001101010110 0000010000</i></p><p><i> 001011000101101001 0000001000</i></p><p><i> 101000011000110101 0000000100</i></p><p><i> 010001110000011011 0000000010</i></p><p><i> 000010100110000111 0000000001</i></p><ul><li><strong>DEC(17, 9)</strong> (low redundancy and reduced overhead DEC) under an <i>aggressive policy </i>(see methodology below):</li></ul><p><i> 111110000 10000000</i></p><p><i> 111001100 01000000</i></p><p><i> 110101010 00100000</i></p><p><i> 101010110 00010000</i></p><p><i> 101101001 00001000</i></p><p><i> 100110101 00000100</i></p><p><i> 100011011 00000010</i></p><p><i> 110000111 00000001</i></p><p>This dataset contains the raw data obtained from:</p><ul><li>running exhaustive fault injection campaigns for increasingly multiple stuck-at faults in the least significant bits of all weights and biases (simultaneously) and for all the images in the workload.</li><li>running statistical fault injection campaigns for single, double, and triple bit-flip faults, randomly targeting the considered locations and images in the workload.</li></ul><h3>Files information</h3><ul><li><i>no_ecc </i>folder: Results obtained for the original (not protected) version of the CNN.<ul><li><i>golden_run.csv</i>: Prediction obtained for all the images considered in the workload in the absence of faults (Golden Run). This is intended to act as oracle to determine the impact of injected faults.</li><li><i>sampling_SBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for single bit-flip faults.</li><li><i>sampling_DBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for double bit-flip faults.</li><li><i>sampling_TBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for triple bit-flip faults.</li><li><i>locating_sensitive_bits </i>folder: Prediction obtained for all the images considered in the workload in presence of stuck-at-0/stuck-at-1 faults that simultaneously target the N least significant bits of all weights and biases. There is one file for each parameter of type of fault and range of targeted bits. Files for bits in the range [11, 0] are not included as they obtain eactly the same results as the Golden Run (faults do not alter the behaviour of the network).</li></ul></li><li><i>sec/classic</i>, <i>sec/conservative</i>, and <i>sec/aggressive</i> folders: They contain the results obtained for the CNN protected by SEC(32, 26), SEC(23, 18), and SEC(13, 9), respectively.<ul><li><i>golden_run.csv</i>: Prediction obtained for all the images considered in the workload in the absence of faults (Golden Run). This is intended to act as oracle to determine the impact of injected faults. It must be noted that this file could be different that the golden_run.csv file for the original version of the CNN, as deploying the ECC in the weights and biases may have affected the behaviour of the network.</li><li><i>sampling_SBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for single bit-flip faults. They should all be tolerated by the definition of the ECC.</li><li><i>sampling_DBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for double bit-flip faults. They could be more harmful than for the unprotected version of the CNN, as the ECC may erroneously flip correct bits.</li></ul></li><li><i>dec/classic</i>, <i>dec/conservative</i>, and <i>dec/aggressive </i>folders: They contain the results obtained for the CNN protected by DEC(32, 21), DEC(28, 18), and DEC(17, 9), respectively.<ul><li><i>golden_run.csv</i>: Prediction obtained for all the images considered in the workload in the absence of faults (Golden Run). This is intended to act as oracle to determine the impact of injected faults. It must be noted that this file could be different that the golden_run.csv file for the original version of the CNN, as deploying the ECC in the weights and biases may have affected the behaviour of the network.</li><li><i>sampling_DBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for double bit-flip faults. They should all be tolerated by the definition of the ECC.</li><li><i>sampling_TBF_10000.csv</i>: Prediction obtained for running 10000 statistical fault injection experiments for triple bit-flip faults. They could be more harmful than for the unprotected version of the CNN, as the ECC may erroneously flip correct bits.</li></ul></li></ul><h3>Methodology information</h3><p>First, the CNN was used to classify all the images of the workload in the absence of faults to get a reference to determine the impact of faults. This is <i>golden_run.csv</i> file.</p><p>To locate non-significant bits in weights and biases, fault injection experiments were executed targeting all elements of all parameters of the CNN using the following procedure:</p><ul><li>The initial mask targeted only the least significant bit</li><li>Until the mask targets all bits of the elements (32 bits as they are single-precision floating point values):<ul><li>Affect the bits (setting them to 0 or 1 in case of stuck-at-0 or stuck-at-1 faults) identified by the mask for all elements of all parameters.</li><li>Classify all the images of the workload in the presence of this fault. The obtained output was stored in a given .csv file.</li><li>Remove the fault from the CNN by restoring the affected bits to its previous value.</li><li>Add the next adjacent bit to the mask, so it targets an additional least significant bit.</li></ul></li></ul><p>The analysis of the obtained results may help in determining which bits can be used to store an ECC:</p><ul><li>which bits never affect the behaviour of the CNN, as the predicted classification is exactly the same than in the absence of faults.</li><li>which bits midly affect the behaviour of the CNN, as although the predicted classifications differ from those in the absence of faults, the accuracy of the network is barely affected.</li><li>which bits greatly affect the behaviour of the CNN, as the accuracy of the network is significantly affected.</li></ul><p>Accordingly, three different policies have been identified for deploying an ECC using these bits:</p><ul><li><strong>Classic policy</strong>: The ECC protects as much bits as possible.</li><li><strong>Conservative policy</strong>: The ECC protects all those bits that may affect the prediction of the network.</li><li><strong>Aggressive policy</strong>: The ECC protects only those bits that significantly affect the accuracy of the network.</li></ul><p>After designing and deploying a single ECC and a double ECC for each of the identified policies, fault injection experiments were executed to verify their behaviour in the presence of faults.</p><p>Single and double ECCs were tested against single and double bit-flip, respectively (all faults should be tolerated,) and double and triple bit-flips, respectively (a correct bit could be erroneously flipped.)</p><p>Due to the heavy computational load of the decoders, statistical injection was used to run the required fault injection campaigns with a sample size (number of experiments) of 10000.</p><p>Each experiment consisted in:</p><ul><li>Randomly selecting the image to process, and the parameter, element, and bits (mask) to be targeted by the fault.</li><li>Affecting the bits (inverting them) identified by the mask.</li><li>Classifying the selected image of the workload in the presence of this fault. The obtained output was stored in a given .csv file.</li><li>Removing the fault from the CNN by restoring the affected bits to its previous value.</li></ul><h3>List of variables (Name : Description (Possible values))</h3><ul><li><strong>IMGID</strong>: Integer number identifying the considered image (1-9999).</li><li><strong>TENSORID</strong>: Integer number identiying the parameter affected by the fault (0 - No fault, 1 - conv1.w, 2 - conv1.b, 3 - conv2.w, 4 - conv2.b, 5 - fc1.w, 6 - fc1.b, 7 - fc2.w, 8 - fc2.b).</li><li><strong>ELEMID</strong>: Integer number identiying the element of the parameter affected by the fault (-1 - No fault, [0-2] - conv1.b, [0-74] - conv1.w, [0-5] - conv2.b, [0-149] - conv2.w, [0-146] - fc1.b, [0-43217] - fc1.w, [0-9] - fc2.b, [0-1469] - fc2.w).</li><li><strong>MASK</strong>: 8-digit hexadecimal number identifying those bits affected by the fault ([00000000 - No fault, FFFFFFFF - all 32 bits faulty]).</li><li><strong>FAULT</strong>: String identiying the type of fault (NF - No fault, BF - bit-flip, S0 - Stuck-at-0, S1 - Stuck-at-1).</li><li><strong>SOFTMAX</strong>: 10 decimal numbers obtained after applying the softmax function to the provided output. They represent the probability of the image of belonging to the corresponding category for classification.</li><li><strong>PRED</strong>: Integer number representing the category predicted for the processed image.</li><li><strong>LABEL</strong>: integer number representing the actual category for the processed image.</li></ul>
Mark loss can strongly bias estimates of demographic rates in multi-state models: a case study with simulated and empirical datasets
<p>This archive contains the empirical data analysed in the paper 'Mark loss can strongly bias estimates of demographic rates in multi-state models: a case study with simulated and empirical datasets' by Touzalin et al. (https://doi.org/10.24072/pci.ecology.100416). The dataset is provided as a .Rdata file ('TLoss_GMdata.Rdata'), and full description of the content is provided in the file 'Readme_TLdata.csv'. All additional details are available in the main text (https://doi.org/10.24072/pci.ecology.100416) or in the supporting information (https://doi.org/10.5281/zenodo.10204538).</p>
Data from: Integrated species distribution models to account for sampling biases and improve range wide occurrence predictions
<p><strong><span>Aim</span></strong></p> <p><span>Species distribution models (SDMs) that integrate presence-only and presence-absence data offer a promising avenue to improve information on species' geographic distributions. The use of such 'integrated SDMs' on a species range-wide extent has been constrained by the often-limited presence-absence data and by the heterogeneous sampling of the presence-only data. Here, we evaluate integrated SDMs for studying species ranges with a novel expert range map-based evaluation. We build a new understanding about how integrated SDMs address issues of estimation accuracy and data deficiency and thereby offer advantages over traditional SDMs.</span></p> <p><strong><span>Location</span></strong></p> <p><span>South and Central America.</span></p> <p><strong><span>Time period</span></strong></p> <p><span>1979-2017.</span></p> <p><strong><span>Major taxa studied</span></strong></p> <p><span>Hummingbirds.</span></p> <p><strong><span>Methods</span></strong></p> <p><span>We build integrated SDMs by linking two observation models – one for each data type – to the same underlying spatial process.</span> <span>We validate SDMs with two schemes: i) cross-validation with presence-absence data and ii) comparison with respect to the species' whole range as defined with IUCN range maps. We also compare models relative to the estimated response curves and compute the association between the benefit of the data integration and the number of presence records in each data set.</span></p> <p><strong><span>Results</span></strong></p> <p><span>The integrated SDM accounting for the spatially varying sampling intensity of the presence-only data was one of the top-performing models in both model validation schemes. Presence-only data alleviated overly large niche estimates, and data integration was beneficial compared to modelling solely presence-only data for species that had few presence points when predicting the species' whole range. On the community level, integrated models improved the species richness prediction.</span></p> <p><strong><span>Main conclusions</span></strong></p> <p><span>Integrated SDMs combining presence-only and presence-absence data are successfully able to borrow strengths from both data types and offer improved predictions of species' ranges. Integrated SDMs can potentially alleviate the impacts of taxonomically and geographically uneven sampling and to leverage the detailed sampling information in presence-absence data.</span></p>
Data from: The role of mutation bias in adaptive molecular evolution: insights from convergent changes in protein function
<p>An underexplored question in evolutionary genetics concerns the extent to which mutational bias in the production of genetic variation influences outcomes and pathways of adaptive molecular evolution. In the genomes of at least some vertebrate taxa, an important form of mutation bias involves changes at CpG dinucleotides: If the DNA nucleotide cytosine (C) is immediately 5' to guanine (G) on the same coding strand, and if the C is methylated, then C→T and G→A mutations occur at an elevated rate relative to mutations at non-CpG sites. Here we examine experimental data from case studies in which it has been possible to identify the causative substitutions that are responsible for adaptive changes in the functional properties of vertebrate hemoglobin (Hb). Specifically, we examine the molecular basis of convergent increases in Hb-O<sub>2</sub> affinity in high-altitude birds. Using a data set of experimentally verified, affinity-enhancing mutations in the Hbs of highland avian taxa, we tested whether causative changes are enriched for mutations at CpG dinucleotides relative to the frequency of CpG mutations among all possible missense mutations. The tests revealed that a disproportionate number of causative amino acid replacements were attributable to CpG mutations, demonstrating that mutation bias can influence outcomes of molecular adaptation.</p>
Fig. 1 in Problematic Biases in the Availability of Molecular Markers in Protists: The Example of the Dinoflagellates
Fig. 1. Number of species of the most speciose dinoflagellate genera (> 11 species per genus). The empty bars represented the number of described species based on Gómez (2012a). The black bars represent the number of species with, at least, one nucleotide sequence available in DDBJ/EMBL/GenBank in January 2013.
Systematic and persistent bias against introduced species
<p>Data used for Pereyra et al 2023 "Systematic and persistent bias against introduced species"</p> <p>The file data_Pereyra_et_al_2023.csv has the dataet, while the file attributes.csv has column descriptions. </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.