DLCSS: A New Similarity Measure for Time Series Data Mining under LCSS

In this research, a new similarity measurement method named Developed Longest Common Subsequence (DLCSS) has been suggested for time series data mining. The main idea of this method is using the logic of the Longest Common Subsequence (LCSS) method and the concept of similarity in time series data. In most studies related to time series data mining, LCSS and Dynamic Time Warping (DTW) had been mentioned as the best and the most usable similarity measurement methods. LCSS has been intrinsically designed to measure the similarity of two sequences of characters, and it was later developed for time series by defining and determining the similarity threshold. Major disadvantage of LCSS is a huge impact of the value of similarity threshold on the quality of time series data mining. In the proposed method, by defining two similarity thresholds and determining their values, LCSS defect was eliminated. To show the effectiveness of DLCSS, it s performance was compared with performance of LCSS and DTW in time series data mining using the Nearest neighbor and K-medoids Clustering techniques on 63 datasets of the UCR datasets. Considering the results, it can be claimed that the performance of DLCSS is better than LCSS and DTW with 98% confidence.

Keywords:

Time Series , Data Mining , Similarity Measurement , Longest Common Subsequence , Dynamic Time Warping , Developed Longest Common Subsequence