Pattern distance of time series based on segmentation by important points
Gao-Zhan Yu, Hong Peng, Qi-Lun Zheng · 2005
In order to analyze the changing trend of time series, a novel method is proposed in this paper, which supports fast searching similar trend sequence in time series. It first segments time series based on a series of perceptually important points, and then converts important point series into piecewise trend sequence (PTS). And a variable-step algorithm for subtrend sequence searching based on PTS is also proposed. The theoretic analysis and simulation indicate that the algorithm has better performance in time and space than classic ones.