Crash Course on Machine Learning for TransportationhEART 2027

PhD Summer School — Monday, 6 September 2027 (09:00–21:30)

Machine learning is increasingly used in transportation research, but many researchers acquire these skills informally — through short tutorials or code shared alongside published papers. This approach is often sufficient to apply a method, but it does not always provide a clear understanding of why a method succeeds or fails on a given problem, which can limit its effective use in original research.
This school is intended to provide a more structured introduction. GMUM, one of the leading machine learning research groups in Poland, has designed each session around problems commonly encountered in transportation research: estimating choice models, processing street-level imagery, modelling route-choice behaviour, and developing surrogate models for computationally expensive simulators. Each session is co-taught by a machine learning researcher and a transportation researcher familiar with the corresponding application.
Participants are expected to have a basic working knowledge of Python and introductory statistics. Each session includes a practical exercise, and participants will leave with code that can be adapted for use in their own research.

• Session 1 — Supervised Learning: Choice Models and Machine Learning
• Session 2 — Computer Vision for Transportation
• Session 3 — Reinforcement Learning: Route Choice as a Multi-Agent Problem
• Session 4 — Graph Neural Networks and Surrogate Models
• Session 5 — Panel Discussion: Publishing at Leading Machine Learning Conferences

Organizers:
Rafał Kucharski's group — http://rafalkucharskilab.pl
GMUM, Jagiellonian University — http://gmum.net
hEART 2027 — http://heart2027.pl