Abstract
Accurately predicting the travel time of each key road in a certain period of time will help the traffic management department to take measures to prevent and reduce traffic congestion. At the same time, it can help to make an optimal travel plan for the traveler based on the dynamic traffic information. Consequently, the utilization efficiency of the load can be improved. RF-DBSCAN, a prediction model based on the random forest (RF) and DBSCAN (Density-Based Spatial Clustering of Applications with Noise), is proposed. After trained using the history traffic datasets, the model can predict the road travel time taking into account the regularity of time series, weather factors, road structures, weekends, and holidays. Experiments are carried out and the results show that the RF-DBSCAN has higher accuracy compared with the traditional random forest and GBDT (Gradient Boosting Decision Tree).
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call