Short-term traffic flow prediction model based on deep learning regression algorithm

Abstract

In view of the problem that the short-term traffic flow prediction under the condition of unsteady traffic flow, such as low precision and over-reliance on large sample historical data, proposing a novel short-term traffic-flow prediction method based on deep learning support vector regression (DL-SVR). A framework of the DL-SVR is built with a restricted Boltzmann machine (RBM) visible inputting layer, which is connected with several intermediate operating networks, and a radial SVR output layer. In addition, a T mutation particle swarm optimisation algorithm is proposed to select the important parameter in DL-SVR. Experimental results show that the mean absolute percentage error (MAPE) and root mean square error (RMSE) of the proposed short-term traffic-flow prediction method are better than other classic algorithms, and the real time also can meet the needs of practical use.