The New Data Structure and Algorithm for Mining Asynchronous Periodic Patterns

• Main Authors: Szu-Chen Lin, 林思甄
• Other Authors: Jieh-Shan Yeh
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/nmbd23

碩士 === 靜宜大學 === 資訊管理學系研究所 === 97 === The periodic pattern mining is to discover valid periodic patterns in a time-related dataset. Previous studies often have considered synchronous periodic patterns. However, asynchronous periodic patterns mining gradual received more and more attention recently. There are many methods that have been proposed for the periodic patterns mining in literature. But those algorithms for dealing with disturbances may neglect some valid patterns. Therefore, the aim of this paper is to offer a more general method of mining asynchronous periodic patterns and to generate all valid periodic patterns. First, we propose OEOP algorithm to discover all kinds of valid segments in each single event sequence. Then, refer to the general model of asynchronous periodic patterns mining proposed by Huang and Chang, we combine these valid segments found by OEOP algorithm into 1-patterns with multiple events, multiple patterns with multiple events and asynchronous periodic patterns. Besides, we implement these algorithms on three real and one synthetic periodic datasets. Then test the relationships of each variable and the efficiency of algorithms. The relationship of the input parameters and the efficiency of algorithms are also examined. The experimental results show that these algorithms have the good performance and scalability.