A combined filtering search for DTW

Due to its high matching precision, DTW is an important similarity measure for time series. However, the high computational complexity restricts its application on mass data sets. To solve the problem, a combined filtering search is proposed, which combines lower-bounding technique and early-abandon strategy, to reduce the redundant computation. The experimental results indicate that the proposed method has the potential to improve the efficiency of similarity search under DTW and guarantee no false dismissals.