Segmentation of Time Series by the Clustering and Genetic Algorithms

This paper proposes a time series segmentation approach by combining the clustering technique, the discrete wavelet transformation and the genetic algorithm to automatically find segments and patterns from a time series. The genetic algorithm is used to find the segmentation points for deriving appropriate patterns. In the fitness evaluation, the proposed algorithm first divides subsequences in a chromosome into k clusters by using the k-means clustering approach. The Euclidean distance is then used to calculate the distance of each subsequence and evaluate a chromosome. The discrete wavelet transformation is also used to adjust the length of the subsequences for comparing their similarity since their length may be different. Experimental results show that the proposed approach can get good effects in finding appropriate segmentation patterns in time series