Study Of Temporal Data Mining Based On Rough Set

In recent years, temporal data emerged in varied fields, and many studies of it had been held. Temporal data includes interesting information and contains some rules of things. To obtain those unknown , available and underlying knowledge which implied in temporal data, we have to study its traits and properties. Some papers had proposed concepts, definition and properties such as temporal type, multi-time granularities, etc. With those definition and properties, we hope to get a better result of data mining. However, it is a problem how to exercise those definition and properties in practice. So this paper proposes temporal relational pattern to realize switching from theory of temporal data to practical application. This paper applies theory of rough set to study temporal data mining.This paper discusses the theory of rough set and actuality of study with respect to temporal data mining, and introduces the contents of study of this paper in chapter 1. It introduces concepts and properties of rough set that relate to this paper's contents of study in chapter 2. The third chapter discusses temporal data and temporal information system with multi-time Granularities. The fourth chapter researches into several algorithms with respect to temporal data mining that based on rough set. The fifth chapter applies our work to analyze the statistic of vehicular circulation. This paper has done the rewarding try in the research of temporal data mining based on rough set. The main achievements contain: 1. Put forward the temporal information system with multi-time granularities, and study its period and properties of multi-time granularities; 2. Bring forward the rules that must be applied when set temporal relational pattern; 3. Develop several algorithms that work in whole procedure of temporal data mining, by which we can apply rough set to obtain period pattern of multi-time granularities on the temporal information system with multi-time granularities.