Publication Date: 12 March 2026
Update Date: 24 March 2026
A study by Dr. Mehmet Taciddin Akçay, Assistant Professor in the Department of Electrical and Electronics Engineering at our University’s Faculty of Engineering, titled “Age-sensitive urban rail passenger demand forecasting and uncertainty-driven anomaly detection using a hybrid SAINT + CatBoost ensemble,” has been published in Scientific Reports, one of the world’s leading academic journals indexed in the SCI Q1 category. The single-authored study examines the urban transportation dynamics of Istanbul.
The research analyzed approximately 721,000 real-world passenger trip records and developed an innovative hybrid artificial intelligence model capable of accurately forecasting metro passenger mobility. Within the study, SAINT, a deep learning architecture, was integrated with CatBoost, a powerful machine learning classification algorithm, to construct an original predictive framework. The proposed model achieved an accuracy rate of 91.94%, demonstrating a notable level of performance in urban transportation forecasting.
As part of the research, the model examined travel behavior across different age groups and generated important analytical insights. While the mobility patterns of the working population were associated with travel frequency, the transportation preferences of older passengers were evaluated through distinct travel patterns. In this way, the behavioral dynamics of urban mobility were revealed through a data-driven analytical perspective.
Published as a single-authored study, the research represents an important example of how complex urban challenges can be addressed through innovative artificial intelligence methodologies. The model presented in the study provides valuable guidance for transportation planners and metro operators in areas such as capacity planning, operational efficiency, and system resilience.
We congratulate our faculty member on this significant achievement and wish him continued success in his academic endeavors.