Time series subsequence matching based on a combination of PIP and clipping

Abstract

Subsequence matching is a non-trivial task in time series data mining. In this paper, we introduce our proposed approach for solving subsequence matching which is based on IPIP, our new method for time series dimensionality reduction. The IPIP method is a combination of PIP (Perceptually Important Points) method and clipping technique in order that the new method not only satisfies the lower bounding condition, but also provides a bit level representation for time series. Furthermore, we can make IPIP indexable by showing that a time series compressed by IPIP can be indexed with the support of Skyline index. Our experiments show that our IPIP method is better than PAA in terms of tightness of lower bound and pruning power, and in subsequence matching, IPIP with Skyline index can perform faster than PAA based on traditional R*-tree. © 2011 Springer Berlin Heidelberg.