An efficient hash-based method for time series motif discovery

Abstract

In this paper, we propose a new efficient method for discovering 1-motifs in large time series data that can perform faster than Random Projection algorithm. The proposed method is based on hashing with three improvement techniques. Experimental results on several benchmark datasets show that our proposed method can discover precise motifs with high accuracy and time efficiency on large time series data.