demo_dataset
收藏资源简介:
该存储库包含一组经过处理的EEG基准数据集,专为模型训练和评估而打包。每个数据集存储在自己的目录中,并遵循一致的基于拆分的HDF5布局,包括训练集、验证集和测试集的HDF5文件,以及Croissant元数据和元数据摘要文件。数据集包含60个不同的EEG数据集,如AD65、ADHD、DEAP等,适用于多种EEG相关任务。数据格式为三维张量(样本数×通道数×时间点),具体参数如采样率、窗口长度和通道数因数据集而异,详细信息记录在metadata_summary.json和croissant.json中。该存储库为处理后的基准集合,非原始数据发布,建议使用前查阅元数据文件以了解任务定义、标签映射和窗口参数。
This repository contains a collection of processed EEG benchmark datasets, packaged specifically for model training and evaluation. Each dataset is stored in its own directory and follows a consistent split-based HDF5 layout, including HDF5 files for training, validation, and test sets, along with Croissant metadata and metadata summary files. The datasets include 60 different EEG datasets such as AD65, ADHD, DEAP, etc., suitable for various EEG-related tasks. The data format is a three-dimensional tensor (number of samples × number of channels × time points), with specific parameters such as sampling rate, window length, and number of channels varying by dataset. Detailed information is recorded in metadata_summary.json and croissant.json. This repository is a processed benchmark collection, not the original data release, and it is recommended to consult the metadata files before use to understand task definitions, label mappings, and window parameters.




