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.
15
datasets available to search
ShareScore release 0.9.0
Dataset results
15 results for “Federated Learning”
MNIST-Federated-Learning
<p>Please find below the descriptions of the three configurations for partitioning <strong>the MNIST Train dataset into 10 clients and the MNIST Train data: </strong><br> </p> <ol> <li><strong>Balanced Distribution:</strong> In the first configuration, the MNIST dataset is partitioned among 10 clients in a balanced manner. This means that the data samples from each class are evenly distributed among the clients. Each client receives a roughly equal number of images from each digit class, ensuring that the distribution of samples across clients is proportional and representative of the overall dataset. <strong> [ Config 1]</strong></li> <li><strong>Heterogeneous Distribution (One Class per Client)</strong>: In the second configuration, the MNIST dataset is partitioned in a heterogeneous manner, where each client is assigned a single digit class exclusively. This means that one client will only receive images of the digit '0', another client will receive images of the digit '1', and so on. In this setup, each client becomes an expert in classifying a specific digit, allowing for specialized training and evaluation. <strong>[ Config 2]</strong></li> <li><strong>Mixed Distribution:</strong> In the third configuration, the MNIST dataset is partitioned using a mixed distribution approach. This means that the data samples from all digit classes are distributed among the 10 clients, but the distribution is not necessarily balanced. The number of samples assigned to each client may vary for different digit classes, resulting in an uneven distribution across the clients. This configuration aims to capture both the overall diversity of the dataset and the varying difficulty levels of classifying different digits. <strong>[ Config 3 ]</strong></li> </ol> <p> </p> <p>Mnist-dataset/<br> ├── config1/<br> │ ├── client-1/<br> │ │ └── data.csv<br> │ ├── client-2/<br> │ │ └── data.csv<br> │ ├── client-3/<br> │ │ └── data.csv<br> │ └── ...<br> ├── config2/<br> │ ├── client-1/<br> │ │ └── data.csv<br> │ ├── client-2/<br> │ │ └── data.csv<br> │ ├── client-3/<br> │ │ └── data.csv<br> │ └── ...<br> ├── config3/<br> │ ├── client-1/<br> │ │ └── data.csv<br> │ ├── client-2/<br> │ │ └── data.csv<br> │ ├── client-3/<br> │ │ └── data.csv<br> │ └── ...<br> └── mnist_test.csv<br> </p> <p>***</p> <p>License: Yann LeCun and Corinna Cortes hold the copyright of MNIST dataset, which is a derivative work from original NIST datasets. MNIST dataset is made available under the terms of the <a href="https://creativecommons.org/licenses/by-sa/3.0/">Creative Commons Attribution-Share Alike 3.0 license.</a></p> <p>***</p>
Federated Learning for Distributed Intrusion Detection Systems in Public Networks - Validation Dataset
<p>This dataset has been meticulously prepared and utilized as a validation set during the evaluation phase of "Meta IDS" to asses the performance of various machine learning models. It is now made available for interested users and researchers who seek a reliable and diverse dataset for training and testing their own custom models.</p> <p>The validation dataset comprises a comprehensive collection of labeled entries, that determines whether the packet type is "malicious" or "benign." It covers complex design patterns that are commonly encountered in real-world applications. The dataset is designed to be representative, encompassing edge and fog layers that are in contact with cloud layer, thereby enabling thorough testing and evaluation of different models. Each sample in the dataset is labeled with the corresponding ground truth, providing a reliable reference for model performance evaluation.</p> <p> </p> <p>To ensure convenient distribution and storage, the dataset has been broken down into three separate batches, each containing a portion of the dataset. This allows for convenient downloading and management of the dataset. The three batches are provided as individual compressed files.</p> <p> </p> <p>In order to extract the data, follow the following instructions:</p> <ul> <li>Download and install bzip2 (if not already installed) from the official website or your package manager.</li> <li>Place the compressed dataset file in a directory of your choice.</li> <li>Open a terminal or command prompt and navigate to the directory where the compressed dataset file is located.</li> <li>Execute the following command to uncompress the dataset: <ul> <li>bzip2 -d filename.bz2</li> </ul> </li> <li>Replace "filename.bz2" with the actual name of the compressed dataset file.</li> </ul> <p>Once uncompressed, you will have access to the dataset in its original format for further exploration, analysis, and model training etc. The total storage required for extraction is approximately 800 GB in total, with the first batch requiring approximately 302 GB, the second batch requiring approximately 203 GB, and the third batch requiring approximately 297 GB of data storage.</p> <p> </p> <p>The first batch contains 1,049,527,992 entries, where as the second batch contains 711,043,331 entries, and for the third and last batch we have 1,029,303,062 entries. The following table provides the feature names along with their explanation and example value once the dataset is extracted.</p> <p> </p> <table align="left"> <thead> <tr> <th scope="col">Feature</th> <th scope="col">Description</th> <th scope="col">Example Value</th> </tr> </thead> <tbody> <tr> <td>ip.src</td> <td>Source IP address in the packet</td> <td>a05d4ecc38da01406c9635ec694917e969622160e728495e3169f62822444e17</td> </tr> <tr> <td>ip.dst</td> <td>Destination IP address in the packet</td> <td>a52db0d87623d8a25d0db324d74f0900deb5ca4ec8ad9f346114db134e040ec5</td> </tr> <tr> <td>frame.time_epoch</td> <td>Epoch time of the frame</td> <td>1676165569.930869</td> </tr> <tr> <td>arp.hw.type</td> <td>Hardware type</td> <td>1</td> </tr> <tr> <td>arp.hw.size</td> <td>Hardware size</td> <td>6</td> </tr> <tr> <td>arp.proto.size</td> <td>Protocol size</td> <td>4</td> </tr> <tr> <td>arp.opcode</td> <td>Opcode</td> <td>2</td> </tr> <tr> <td>data.len</td> <td>Length</td> <td>2713</td> </tr> <tr> <td>eth.dst.lg</td> <td>Destination LG bit</td> <td>1</td> </tr> <tr> <td>eth.dst.ig</td> <td>Destination IG bit</td> <td>1</td> </tr> <tr> <td>eth.src.lg</td> <td>Source LG bit</td> <td>1</td> </tr> <tr> <td>eth.src.ig</td> <td>Source IG bit</td> <td>1</td> </tr> <tr> <td>frame.offset_shift</td> <td>Time shift for this packet</td> <td>0</td> </tr> <tr> <td>frame.len</td> <td>frame length on the wire</td> <td>1208</td> </tr> <tr> <td>frame.cap_len</td> <td>Frame length stored into the capture file</td> <td>215</td> </tr> <tr> <td>frame.marked</td> <td>Frame is marked</td> <td>0</td> </tr> <tr> <td>frame.ignored</td> <td>Frame is ignored</td> <td>0</td> </tr> <tr> <td>frame.encap_type</td> <td>Encapsulation type</td> <td>1</td> </tr> <tr> <td>gre</td> <td>Generic Routing Encapsulation</td> <td>'Generic Routing<br> Encapsulation (IP)’</td> </tr> <tr> <td>ip.version</td> <td>Version</td> <td>6</td> </tr> <tr> <td>ip.hdr_len</td> <td>Header length</td> <td>24</td> </tr> <tr> <td>ip.dsfield.dscp</td> <td>Differentiated Services<br> Codepoint</td> <td>56</td> </tr> <tr> <td>ip.dsfield.ecn</td> <td>Explicit Congestion<br> Notification</td> <td>2</td> </tr> <tr> <td>ip.len</td> <td>Total length</td> <td>614</td> </tr> <tr> <td>ip.flags.rb</td> <td>Reserved bit</td> <td>0</td> </tr> <tr> <td>ip.flags.df</td> <td>Don't fragment</td> <td>1</td> </tr> <tr> <td>ip.flags.mf</td> <td>More fragments</td> <td>0</td> </tr> <tr> <td>ip.frag_offset</td> <td>Fragment offset</td> <td>0</td> </tr> <tr> <td>ip.ttl</td> <td>Time to live</td> <td>31</td> </tr> <tr> <td>ip.proto</td> <td>Protocol</td> <td>47</td> </tr> <tr> <td>ip.checksum.status</td> <td>Header checksum status</td> <td>2</td> </tr> <tr> <td>tcp.srcport</td> <td>TCP source port</td> <td>53425</td> </tr> <tr> <td>tcp.flags</td> <td>Flags</td> <td>0x00000098</td> </tr> <tr> <td>tcp.flags.ns</td> <td>Nonce</td> <td>0</td> </tr> <tr> <td>tcp.flags.cwr</td> <td>Congestion Window Reduced<br> (CWR)</td> <td>1</td> </tr> <tr> <td>udp.srcport</td> <td>UDP source port</td> <td>64413</td> </tr> <tr> <td>udp.dstport</td> <td>UDP destination port</td> <td>54087</td> </tr> <tr> <td>udp.stream</td> <td>Stream index</td> <td>1345</td> </tr> <tr> <td>udp.length</td> <td>Length</td> <td>225</td> </tr> <tr> <td>udp.checksum.status</td> <td>Checksum status</td> <td>3</td> </tr> <tr> <td>packet_type</td> <td>Type of the packet which is either "benign" or "malicious"</td> <td>0</td> </tr> </tbody> </table> <p>Furthermore, in compliance with the GDPR and to ensure the privacy of individuals, all IP addresses present in the dataset have been anonymized through hashing. This anonymization process helps protect the identity of individuals while preserving the integrity and utility of the dataset for research and model development purposes.</p> <p> </p> <p>Please note that while the dataset provides valuable insights and a solid foundation for machine learning tasks, it is not a substitute for extensive real-world data collection. However, it serves as a valuable resource for researchers, practitioners, and enthusiasts in the machine learning community, offering a compliant and anonymized dataset for developing and validating custom models in a specific problem domain.</p> <p> </p> <p>By leveraging the validation dataset for machine learning model evaluation and custom model training, users can accelerate their research and development efforts, building upon the knowledge gained from my thesis while contributing to the advancement of the field.</p>
Information Leaks in Federated Learning
<p>With the surge in data collection and analytics, concerns are raised with regards to the privacy of the individuals represented by the data. In settings where the data is distributed over several data holders, federated learning offers an alternative to learn from the data without the need to centralize it in the first place. This is achieved by exchanging only model parameters learned locally at each data holder. This greatly limits the amount of data to be transferred, reduces the impact of data breaches, and helps to preserve the individual’s privacy. Federated learning thus becomes a viable alternative in IoT and Edge Computing settings, especially if the data collected is sensitive.</p> <p>However, risks for data or information leaks still persist, if information can be inferred from the models exchanged. This can e.g. be in the form of membership inference attacks. In this paper, we investigate how successful such attacks are in the setting of sequential federated learning. The cyclic nature of model learning and exchange might enable attackers with more information to observe the dynamics of the learning process, and thus perform a more powerful attack.</p> <p> </p>
Outputs of the paper "Blockchain-enabled Server-less Federated Learning"
<p>This repository contains the different outputs generated for the paper "Blockchain-enabled Server-less Federated Learning", submitted to Computer Networks (COMNET) journal. Two types of outputs are provided:</p> <ol> <li><strong>Blockchain queue simulator (output_queue_simulator):</strong> results of the simulations done in the batch-service queue simulator (https://github.com/fwilhelmi/batch_service_queue_simulator) to characterize the queue latency of blockchain applications.</li> <li><strong>Tensorflow (output_tensorflow):</strong> results of the simulations done in Tensorflow Federated (TFF), resulting from the application of different models to the federated EMNIST dataset.</li> </ol> <p>Each folder also includes the scripts used to execute the corresponding simulations. For more details, see the repository in https://github.com/fwilhelmi/blockchain_enabled_federated_learning</p>
DVB-T Spectrum Sensing for Federated Learning
<p>These files contain measurement results achieved in the following scenario:<br>Rohde&Schwarz SMBV100A signal generator transmits OFDM signal (center frequency 2000.0 MHz) with different amplifier gain.<br>A single PC with GNU Radio and connected USRP receives a signal with a central frequency of 2000 MHz (filtered 8 MHz bandwidth). <br><br>Received is set in six different positions (location of sensors presented on the figure "position.png") - creating 6 files - and collects data to detect the presence of a transmitted signal. Data that can be found in the files is stored in CSV format to be easily analyzed in ML models. </p> <p>Data collected during this experiment contains:<br>- column with average received power in the analyzed channel (in mW)<br>- column with autocorrelation value<br>- column with decision metric (calculated as maximum eigenvalue divided by minimum eigenvalue - look below)<br>- 8 columns with eigenvalues of the Pearson correlation matrix (8 subsequent vectors of 32768 samples)<br>- column with information about signal transmission (label; 0 - noise, 1 - signal transmitted)</p>
Federated Learning Games for Reconfigurable Intelligent Surfaces via Causal Representations: SW and Data
<p>This upload contains the main simulation code and related datasets used in the conference paper entitled <a href="https://ieeexplore.ieee.org/document/10437657" target="_blank" rel="nofollow noreferrer noopener">Federated Learning Games for Reconfigurable Intelligent Surfaces via Causal Representations</a>, which was presented at <a href="https://globecom2023.ieee-globecom.org/" target="_blank" rel="nofollow noreferrer noopener">IEEE GLOBECOM 2023</a>.</p>
A federated learning framework based on transfer learning and knowledge distillation for targeted advertising-Click-Through Rate Prediction Dataset
<p>https://www.kaggle.com/c/avazu-ctr-prediction</p>
Federated learning dataset: A case study of vibration analysis for desktop 3D printers
<p>This dataset contains the acceleration data collected from six low-cost 3D printers (same make and model) air-printing cubes with different printing speeds. This dataset can be used as a case study for federated learning, personalized learning, etc.</p>
A multi-center MRI dataset for bladder cancer and baseline evaluations of federated learning in its clinical application
Open the record for dataset details and reuse information.
A federated learning framework based on transfer learning and knowledge distillation for targeted advertising-Ad Display/Click Data on Taobao.com dataset
<p>https://tianchi.aliyun.com/dataset/56</p>
Visualization Engineering Platform for TCM Pulse Diagnosis - Pulse Diagnosis Based on Federated Learning to Diagnose Slippery and Choppy and Other Pulses Waveform Image Features to Assist in the Study
ClinicalTrials.gov study NCT05630248. IPD Sharing: NO. Countries: 1. Publications: 0.
Hypertrophic Cardiomyopathy Federated Learning Implementation Platform
ClinicalTrials.gov study NCT06461468. IPD Sharing: NO. Countries: 1. Publications: 0.
FedDL: Personalized Federated Deep Learning for Enhanced Detection and Classification of Diabetic Retinopathy
Open the record for dataset details and reuse information.
Federal Learning Algorithm for an Intelligent Insulin Decision System for Dynamic Glucose Control in Type 2 Diabetic Patients
ClinicalTrials.gov study NCT06434623. IPD Sharing: UNDECIDED. Countries: 0. Publications: 0.
Federated Learning for 5G SS
<p>Part of the data for Federated Learning-based 5G spectrum sensing</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.