Discovery of asynchronous periodic patterns with multiple granularities
挖掘多粒度时间下异步周期的模式

An algorithm for efficient mining of asynchronous periodic patterns with multiple granularities by exploring some interesting properties related to asynchronous periodicity was proposed. The algorithm generated all valid time section at first, and then found the longest valid time segment by using valid time section. The performance study shows that the proposed method is stable and efficient.