Mining Relationship Patterns in Multiple Time Series Based on IRST

Time series are an important type of data, and discovering frequent patterns in time series is a basic task to predicate the changing trend In this paper, an algorithm for the discovery of frequent patterns in multiple time series is proposed In this algorithm, firstly the states relationship between in time series is represented in Allen temporal logic, and then a sliding window is used to examine the order or occur rence relationship of states and obtain a particular sequence On the basis of the sequence, a frequent pattern method called inter related successive trees（IRST）, is developed to find the frequent relationship patterns from multiple time series Compared with the previous methods, the method is more simple, flexible, efficient and more applied value