Akademik Veri Yönetim Sistemi
2nd International Conference on Science, Engineering Management and Information Technology, SEMIT 2023, Ankara, Türkiye, 14 - 15 Eylül 2023, cilt.2198 CCIS, ss.237-252, (Tam Metin Bildiri)
With the remarkable development of Intelligent Transportation Systems in recent years, the easy collection of traffic information and various vehicle-related data has been made possible. The availability of probe information allows for the acquisition of more extensive traffic data in addition to observation information. In this paper, the traffic flow prediction method on urban roads using probe information is considered. Accurate and real-time traffic information is deemed indispensable for the deployment of high-performance intelligent transportation systems. Traffic flow, being a complicated phenomenon, can be expressed in terms of its characteristics without prior knowledge of site-specific features, thanks to the application of deep learning, which automatically acquires feature quantities. Consequently, a traffic flow prediction model utilizing deep learning is investigated in this research. Additionally, a comparison is made with other traffic flow prediction methods. The research aims to improve prediction accuracy and contribute to the advancement of more efficient intelligent transportation systems.