WHY DID TIME SERIES SUBSEQUENCE CLUSTERING DISGUISE US: THE OVERALL ANALYSIS
Shitong Wang, KORRIS F. L. CHUNG, Zhaohong Deng · International Journal of Information Acquisition · 2005
In recent years, time series subsequence clustering is often reported to be very efficient in the data mining community. However, in this article, a negative justification will be given. The overall theoretical analysis and experiments on counterexamples indicate that time series subsequence clustering almost always does not yield any useful result.