Seminar on Traffic Forecasting with Neural Networks (FSS 2025)

In this seminar, we study different approaches for traffic forecasting. Traffic forecasting is essential for improving transportation systems, reducing congestion, and enhancing safety by predicting traffic flow patterns. It helps city planners and authorities make informed decisions to create more efficient and sustainable urban mobility solutions.

Organization

Goals

In this seminar, you will

  • Explore relevant research topics/papers in the realm of traffic forecasting
  • Give an overview of your topic to your peers in a presentation before the easter break
  • Write a final research report, tackling a practical challenge with real-world data provided by the city of Mannheim with the approach you presented before

Schedule

tbd (first meeting will be scheduled via Doodle)

Requirements

  • Good reading and presentation skills in English
  • Proficiency in LaTeX
  • Some programming experience

Evaluation

The grade consists of two parts:

  • 25% are assigned for an individual presentation of a research topic
  • 75% are given for a final research report

Topics and Literature

  • Jiang, Weiwei, and Jiayun Luo. “Graph neural network for traffic forecasting: A survey.” Expert systems with applications 207 (2022): 117921.
  1. Multivariate Time-Series Forecasting
    • Salinas et al., “DeepAR: Probabilistic forecasting with autoregressive recurrent networks”, IJF 2020
    • Hollmann, Noah, et al. “Accurate predictions on small data with a tabular foundation model.” Nature 637.8045 (2025)
    • Shao, Zezhi, et al. “Spatial-temporal identity: A simple yet effective baseline for multivariate time series forecasting.” CIKM. 2022.
  2. Graph-based Traffic Forecasting
    • Cini, Marisca, Zambon, and Alippi, “Taming Local Effects in Graph-based Spatiotemporal Forecasting”, NeurIPS 2023
    • Wu et al., “TraverseNet: Unifying Space and Time in Message Passing for Traffic Forecasting”, TNNLS 2022
  3. Data Imputation for Traffic Forecasting
    • Cini, Andrea, Ivan Marisca, and Cesare Alippi. “Filling the G_ap_s: Multivariate Time Series Imputation by Graph Neural Networks.”, ICLR 2022