Time series clustering of metropolitan railway traffic using dynamic time warping
HyeonA Park1 · Jong Hwa Na2
1Korea Statistical Information Institute
2Department of Information and Statistics, Chungbuk National University
2Department of Information and Statistics, Chungbuk National University
Correspondence to: 1 Researcher, KOSII, 18, Dunsan-daero 117, Seo-gu, Daejeon 35203, Korea.
2 Professor, Department of Information & Statistics, Chungbuk National University, Chungbuk 28644, Korea. E-mail: cherin@cbnu.ac.kr
2 Professor, Department of Information & Statistics, Chungbuk National University, Chungbuk 28644, Korea. E-mail: cherin@cbnu.ac.kr
Received August 2, 2022; Revised September 2, 2022; Accepted September 13, 2022.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
- Abstract
- In this study, time series cluster analysis was performed by applying the DTW algorithm to the metropolitan railroad traffic data. Unlike the Euclidean distance used for clustering of general multivariate data, DTW is well applied in the case of time stretching and distortion, and therefore it is usefully used as a measure of the distance between two time series. Six internal CVI measures were used to evaluate the optimal number of clusters and performance in cluster analysis based on the traffic data of metropolitan railways. As a result, it was found that the degree of cohesion and separation of the cluster using DTW was the greatest compared to the classical cluster. As a result of the clustering, three clusters were formed, and as a result of comparing the central patterns, distinct clustering results were confirmed.
- Keywords : Clustering validation index, dynamic time warping, silhouette index, time series clustering.
Full Text(PDF) Free
