Anomaly Detection in Traffic Networks

Along with the growing traffic networks, the control of traffic networks has become increasingly difficult. The detection of anomaly is very important in ensuring this monitoring. With anomaly detection approaches in traffic networks, events are detected early and quick intervention is provided. This saves time and money. There are classification, clustering, statistical, etc. approaches in anomaly detection for problems in the literature. Support vector machines, Bayesian networks, fuzzy logic, genetic algorithms, etc. are main approaches that used to detect anomalies. In this study, a decision tree algorithm is proposed to detect anomaly in traffic networks. The proposed method has been tested using the British Colombia traffic data set. It is seen that some abnormal events in the traffic network can be detected by using the proposed method