XTSC-Bench
收藏资源简介:
XTSC-Bench是由FZI Research Center for Information Technology创建的数据集,专注于时间序列分类的解释性评估。该数据集包含120个合成数据集,分为单变量和多变量两种类型,每个数据集包含50个时间步长。数据集通过模拟不同的时序过程和信息特征生成,旨在为评估解释方法提供标准化的数据基础。XTSC-Bench的应用领域主要集中在提高时间序列分类模型的解释性和可靠性,特别是在处理复杂的多元时间序列数据时。
XTSC-Bench is a dataset created by the FZI Research Center for Information Technology, focusing on interpretable evaluation for time series classification. It consists of 120 synthetic datasets, categorized into univariate and multivariate types, with each containing 50 time steps. Generated by simulating various temporal processes and informative features, this dataset aims to provide a standardized data basis for evaluating interpretability methods. The application scenarios of XTSC-Bench mainly center on enhancing the interpretability and reliability of time series classification models, particularly when handling complex multivariate time series data.




