Combination Prediction Model of Traffic Flow Based on Rough Set Theory

The prediction of traffic flow plays an important part in intelligent transportation system. Due to the nonlinear and stochastic characteristic of traffic flow, it is difficult to predict traffic flow accurately. In order to improve the prediction precision, a combination prediction model based on rough set and knowledge entropy is proposed. The relative data model between prediction object and prediction model, and the decision table are established by means of converting continuous attribute values into discrete attribute values. Then the weight coefficients of the combination prediction model are determined by evaluating significance of every single prediction model with rough set and knowledge entropy theory. The proposed approach overcomes the limitation of the single prediction model, and makes the determination of weight coefficients more objective. Simulation results show the proposed combination prediction model outperforms any of the single prediction models.