Online travel time prediction based on boosting

Travel time prediction is a very important problem in intelligent transportation system research. We examine the use of boosting, a machine learning technique in travel time prediction, and combine boosting and neural network models to increase prediction accuracy. In addition, quality of service (QoS) factors such as bandwidth play an important role in travel time prediction, so we also explore the relationship between the accuracy of travel time prediction and the frequency of traffic data collection with the long term goal of minimizing bandwidth consumption. Finding a lower bound on the data collection frequency is also an important preliminary step for the boosting-based approach. To evaluate the effectiveness of the proposed algorithm, we conducted three sets of experiments that show the boosting neural network approach outperforms other predictors.