Data collected for A review on EEG-based music datasets

This narrative review has gathered and contrasted the distinct attributes of the openly accessible EEG datasets they are commonly used to develop (affective) Brain-computer Music related algorithms. We explored the strengths and drawbacks inherent in the attributes of the EEG datasets. Drawing from our investigation, we pinpointed 19 characteristics that make the music-based EEG datasets unique. We also briefly looked into how specific characteristics of the datasets that are openly accessible impact the performance and results of a study and influence the selection of machine learning techniques and preprocessing steps required to develop methods to build Brain-computer music interface algorithms. In conclusion, this study provides a guideline for choosing publicly available music-based EEG datasets for musicians and scientists working to develop reproducible, generalizable, and effective brain-computer music algorithms.本叙述性综述收集并对比了当前常用于开发（情感）脑机音乐相关算法的公开可获取脑电图（Electroencephalogram，EEG）数据集的不同特性。本研究探讨了上述脑电图数据集特性中固有的优势与局限。基于本次调研，我们明确了19项赋予基于音乐的脑电图数据集独特性的特征。此外，本研究还简要分析了公开可获取数据集的特定特性如何影响研究的性能与结果，以及其对开发脑机音乐接口（Brain-computer Music Interface）算法所需的机器学习技术与预处理步骤选取的影响。综上，本研究为致力于开发可复现、可泛化且高效的脑机音乐算法的音乐从业者与科研人员，提供了一套公开可用的基于音乐的脑电图数据集选型指南。