A supervised learning approach to estimating the urban traffic state based on floating car data

Abstract

Intelligent transportation systems aim to increase the efficiency of mobility through the use of new information and communication technologies. This paper presents a supervised learning approach to estimate the traffic state in the urban road networks. The data used in this work were collected from the GPS sensors installed on the connected vehicles. We validate the proposed model using the k-fold cross-validation method and we evaluate the prediction accuracy with the root mean square error and the mean absolute error metrics. The proposed model was evaluated by numerical simulation. The results obtained by estimating the mean travel time in the road network using the proposed method showed good accuracy with a ratio of the connected vehicles not exceeding 15%.