Evaluating the Effect of Time Series Segmentation on STARIMA-Based Traffic Prediction Model

摘要: As the interest for developing intelligent transportation systems increases, necessity effective traffic prediction techniques becomes profound. Urban short-term has proven to be an interesting yet challenging task. The goal is forecast values of appropriate descriptors such as average travel time or speed, one more intervals in future. In this paper a novel and efficient approach based on series analysis provided. Our idea split into segments (that represent different trends) use models series. proposed method was evaluated using historical GPS data from city Berlin, Germany covering total period two weeks. results show smaller error, terms time, with respect basic relevant literature.