Hierarchical Matching for Multivariate Time Series Based on Moving Average

• 作 者: YanWANG;QianqianMA;MengHAN;
• 机 构: College of Computer and Communication,Lanzhou University of Technology,Lanzhou,China
• 出 处: 第四届信息电子与计算机工程国际会议(The 4th International Conference on Inform
• 发表日期: 2012年期

　　Since the length and the dimension of multivariate time series can be vary large, the similarity calculation is often computationally expensive.This paper proposes the hierarchical algorithm for the multivariate time series matching based on moving average.Firstly, the rough matching based on moving average is performed to select candidate sequences according to the corollary of distance reduction theorem.Then, the second matching is performed on the basis of representing the selected candidate sequences with another mode.Finally, we get similar sequences by calculating the Euclidean distance.The experiment demonstrates that the search space is pruned gradually in a short time, the sequences which similar to the given query sequence are found efficiently, and the global feature and the local feature can be kept at the same time.