Dynamic Data Streams

该数据集包含了十个来自非平稳档案的动态数据流，用于验证针对数据流变化所提出的方法的适应性。这些数据流包括具有不同特性的合成数据流和一个基于按键动态的真实数据流。具体来说，这些数据流涵盖了从两到四个类别、维度从2到10的合成示例，以及一个基于多次输入的打字节奏来识别用户的按键数据流。这十个数据流在类别和维度上各不相同，其任务是对数据流进行分类。

This dataset contains ten dynamic data streams from non-stationary archives, developed to verify the adaptability of proposed methods for handling data stream changes. These streams include synthetic data streams with diverse characteristics and one real-world data stream based on keystroke dynamics. Specifically, the synthetic streams cover cases with 2 to 4 categories and feature dimensions ranging from 2 to 10, while the real-world stream is a keystroke data stream for user identification built upon typing rhythms extracted from multiple input attempts. All ten data streams differ in terms of category count and feature dimensionality, and the core task of this dataset is data stream classification.