基于依赖度的时序数据的特征选择方法

A Feature Selection Method for Time Series Data Based on Dependence

摘要: 随着大数据时代的不断发展，时序数据广泛存在于生活的各个领域。但现有的信息系统无法存放时序数据或者分类准确率较低。因此，本文构建时序模糊信息表，引入模糊相似关系，提出可以存放时序数据的时序模糊决策粗糙集模型，并研究其性质，给出基于时序依赖度的特征选择方法。

Abstract: With the continuous development of the era of big data, time series data exists in all fields of life. However, the existing information system cannot store time series data or has a low classification accuracy. Therefore, this paper constructs a time series fuzzy information table, introduces fuzzy similarity relations, proposes a rough set model of time series fuzzy decisions that can store time series data, studies its properties, and gives a feature selection method based on time series dependence.