Trajectory pattern mining via clustering based on similarity function for transportation surveillance

• Main Authors: Choong, Mei Yeen, Chin, Renee Ka Yin, Yeo Kiam Beng @ Abdul Noor, Teo, Kenneth Tze Kin
• Format: Article
• Language: English; English
• Published: United Kingdom Simulation Society 2016
• Subjects: QA75.5-76.95 Electronic computers. Computer science; TA1001-1280 Transportation engineering
• Online Access: https://eprints.ums.edu.my/id/eprint/32418/1/Trajectory%20pattern%20mining%20via%20clustering%20based%20on%20similarity%20function%20for%20transportation%20surveillance.pdf; https://eprints.ums.edu.my/id/eprint/32418/3/Trajectory%20pattern%20mining%20via%20clustering%20based%20on%20similarity%20function%20for%20transportation%20surveillance%20_ABSTRACT.pdf; https://eprints.ums.edu.my/id/eprint/32418/; https://ijssst.info/Vol-17/No-34/paper19.pdf; http://dx.doi.org/10.5013/IJSSST.a.17.34.19
• Institution: Universiti Malaysia Sabah

Recently, surveillance on moving vehicles for traffic flow monitoring has emerging in rapid rate. A comprehensive traffic data, that is vehicle trajectory, is selected as reliable data for discovering the underlying pattern via trajectory mining. As the task of monitoring moving vehicles via vehicle trajectory dataset can be tedious, researchers are keen to provide solutions that reducing the tedious task performed by the traffic operators. One of the solutions is to group the vehicle trajectory data according to the shape of the patterns. This grouping task is called as clustering. Each of the clusters formed represents a pattern. In this paper, the analysis of the implemented clustering algorithm on the trajectory data with similarity function is presented. Discussion on the issues concerning the trajectory clustering is also presented.