Real-time Traffic Prediction Using AOSVR and Cloud Model

Abstract

Accuracy and time efficiency in prediction are couple of contradictions to be hard to resolve for real-time traffic information prediction. In order to improve time efficiency of prediction, we develop a real-time traffic information prediction model on the basis of Accurate On-line Support Vector Regression (AOSVR) in this paper, and a simplified computing method of sigmoid kernel based on cloud model is also proposed. Experiments are given to verify the performance of the developed predicting model, and the results obtained show that it obviously improves the time efficiency of predicting in spite of small decrease in precision due to simplifying computing of sigmoid kernel.