Highway Toll Forecasting Model

Abstract

Since the complexity of artificial design features in model training, the toll data features cannot be used reasonably and efficiently. Toll prediction of one single station based on toll historical data would ignore the interaction between stations in the highway network. Therefore, this paper constructs a highway toll forecast model, named DBN-MSVR combining deep belief network and multi-task learning for multi-station toll forecasting. This model uses the optimized deep belief network to perform feature learning on toll data, and combines multi-task learning and support vector regression on the top layer of the deep belief network to predict tolls. Experiments show that the DBN-MSVR toll prediction model has higher prediction accuracy than traditional methods.