Mining Segment-Wise Periodic Patterns with Unknown Periods

摘要：

The search for segment-wise periodic patterns in time series data is an active topic of research. Given the fact that rarely a real world dataset is perfectly periodic, this paper approaches this problem in terms of data mining, trying to discover segment-wise periodic patterns with low periodic support percentage (PSP) in lime series databases, when no period length is known in advance. And the noisy items are randomly generated. So the partial periodic pattern without noisy is a especial example of segment-wise periodic patterns. In existing time series data, utilizing string matching (SM) and shift wavelet tree (SWT). These experiments show that the algorithm is effective to find cyclic association rules and segment-wise periodic signals.

关键词： Shift wavelet tree string matching segment-wise periodic cyclic association rule

作者: Xiaofang You Zhenxing Qin Shichao Zhang

作者单位: Wireless Services Division Tencent Technology Company Limited Shenzhen, China Faculty of EIT, UTS, Australia Faculty of EIT, UTS, Australia State Key Laboratory of Novel Software Technology Nanjing University

会议类型: 国际会议

会议名称: 2010 International Conference on Information Security and Artificial Intelligence(2010年信息安全与人工智能国际会议 ISAI 2010)

会议地点: 成都

会议语种:英文

页码: 1434-1438

在线出版日期: 2010-12-17（万方平台首次上网日期，不代表论文的发表时间）