New similarity measure method on time series
Zhang Hai-ta · Jisuanji gongcheng yu sheji · 2014
Similarity measure on time series is the basis of time series data mining,a new similarity measure method on time series—SMVT by describing time series with symbol based on variation trends and defining trend distance.The variation trends similarity of time series are measured with this method,which is concise and straightforward,trends description scientific,and reflects the overall variation trends of time series well.Applying the new similarity method for clustering computation of time series,the experimental results show that the method can measure the similarity of time series effectively from the perspective of variation trends,which explains its well practicability and good application prospect.