Open PhD Position: Model interpretation and data-centric modeling for advanced traffic prediction (MINDMAP) project (TSE 126)

Description of the Chair

The Chair of Transportation Systems Engineering (TSE) undertakes research in the transportation field with a specific focus on modeling and simulating transportation systems, implementing data science and data analytics in transport and human factors analysis, and machine learning and deep learning. The TSE chair researches multimodal and unimodal freight and passenger transport demand and supply modeling, allowing for contributions to the optimization, calibration, and validation of transport models. In this direction, the application of big data acquisition and analysis and the use of flexible data-driven models are examined. The TSE chair also contributes to analyzing human factors in transport-related fields such as road safety modeling, behavioral economics applications, and modeling of factors that affect user engagement in transportation systems.

Project description

Research in traffic prediction focuses on managing evolving demand and supply dynamics. Deep learning, particularly Graph Neural Networks (GNNs), is proven effective, emphasizing model accuracy with pre-processed datasets. Transfer learning shows promise for applying pre-trained models to new areas, though challenges arise due to real-world data complexities. The MINDMAP project explores data-efficient and interpretable transfer of traffic prediction models, addressing interpretability gaps in GNNs. Data-centric methods transform raw data systematically, while interpretability methods shed light on transferability of demand and supply elements. The project aims to integrate these approaches into a real-time network-wide traffic prediction prototype/dashboard, bridging the research-practice gap and advancing the state-of-the-art in traffic prediction. 

 

The three groups collaborating on this international project are the chair of Transportation Systems Engineering (TSE) of TUM (Prof. Dr. Constantinos Antoniou), the group of Prof. Dr. Jiwon Kim from the University of Queensland, Australia, and the Data Analytics and Machine Learning (DAML) group of TUM (Prof. Dr. Stephan Günnemann). The research team will approach the ambitious research goals set out in this proposal from the perspectives of transferability, interpretability, and data-centric perspectives, using distinct but complementary methodologies and access to different sets of collaborators and resources while striving for a common goal. 

Requirements