In order to solve the problem that the performance of traditional Automatic Incident Detection (AID) detector was restricted due to the imbalanced training set, a SVDD (Support Vector Domain Description) algorithm was brought in to detect traffic incidents. Using actual traffic data from the America I-880 freeway, the proposed SVDD-AID algorithm was trained and its performance was tested. The experimental results show that the model parameters should be chosen carefully because they have great effects on the classification result, and the SVDD-AID algorithm can obtain desirable detection performance.
展开▼