摘要
本文对基于片断模式的多时间序列关联分析进行了研究,提出了一种分析方法。这一方法是,首先通过聚类找出在时间序列中频繁出现的片断模式,然后将找到的片断模式作为模板,对时间序列进行跨事务关联分析。我们采用中国证券市场1997～2001年的数据为测试数据集,对我们提出的算法进行了测试。测试结果表明,我们的算法是有效的。
In this paper, it is studied that segment-based multiple time series association rules analysis, and an approach is presented. The approach is that, firstly finding the frequent segment pattern in the time series by means of clustering, then using the found segment patterns as templates to do the inter-transactional association analysis. We use the data of Chinese stock market from 1997 to 2001 as the test data set to test our approach. The experimental resuits show that the approach is effective.
出处
《计算机科学》
CSCD
北大核心
2006年第1期232-235,共4页
Computer Science
