TIDMAD

TIDMAD是由麻省理工学院物理系创建的用于发现暗物质的超长时序数据集。该数据集包含三个部分：训练数据、验证数据和科学数据，总计约80.23亿样本，用于支持AI算法提取暗物质信号。数据集通过ABRACADABRA实验收集，该实验每秒产生1000万样本，旨在通过AI去噪技术提高暗物质探测的灵敏度。TIDMAD的应用领域主要集中在物理学的基础研究，特别是暗物质的探测和理解。

TIDMAD is an ultra-long time-series dataset created by the Department of Physics, Massachusetts Institute of Technology (MIT) for dark matter discovery. This dataset includes three parts: training data, validation data and scientific data, with a total of approximately 8.023 billion samples, designed to support AI algorithms in extracting dark matter signals. Collected through the ABRACADABRA experiment which generates 10 million samples per second, this dataset aims to improve the detection sensitivity of dark matter via AI denoising technologies. The application fields of TIDMAD mainly focus on basic physics research, especially the detection and understanding of dark matter.