Finding Discordant Subsequence in Multivariate Time Series

Discordant subsequence in multivariate time series (MTS) is the subsequence that is least similar to all other MTS subsequences. In this paper, an algorithm of finding discordant subsequence in MTS, based on solving set, is proposed. Subsequences can be extracted by use of a sliding window. An extended Frobenius norm is used to compute the distance between MTS subsequences. The time complexity of the algorithm is subquadratic in the length of the MTS. We conduct experiments on two real-world datasets, stock market dataset and BCI (Brain Computer Interface) dataset. The experiment results show the efficiency and effectiveness of the algorithm.