Highway Traffic Forecasting By Support Vector Regression Model With Tabu Search Algorithms

Accurate forecasting of inter-urban traffic flow has been one of most important issues globally in the research on road traffic congestion. The information of inter-urban traffic with a cyclic data pattern presents a more challenging situation. This investigation presents a short-term traffic forecasting model which combines the support vector regression model with Tabu search algorithms (SVRTA) to forecast inter-urban traffic flow. The Tabu search algorithms (TA) are used to determine the three parame-ters of support vector regression (SVR) models. Additionally, a numerical ex-ample of traffic flow values from northern Taiwan is used to elucidate the fore-casting performance of the proposed SVRACO model. The simulation results indicate that the proposed model yields more accurate forecasting results than the seasonal autoregressive integrated moving average (SARIMA) time-series model. Therefore, the SVRTA model is able to capture the cyclic data pattern and therefore a promising alternative in forecasting urban traffic