Communication Dans Un Congrès Année : 2025

Machine Learning Surrogates for Optimizing Transportation Policies with Agent-Based Models

Résumé

This study aims to evaluate the feasibility of using Graph Neural Networks as surrogates for agent-based models in transport planning, with the long-term goal of enabling simulation-based optimization through agent-based simulations. To the authors’ knowledge, only Narayanan et al. has provided a proof-of-concept, replicating the 4-step model with GNNs on networks of 15 to 80 nodes. Our study, by contrast, trains a GNN using MATSim simulation data for a Paris case study with 30,000 nodes. The results indicate that the trained model can successfully predict outcomes across different policy implementations.

Fichier non déposé

Dates et versions

hal-04892998 , version 1 (17-01-2025)

Identifiants

  • HAL Id : hal-04892998 , version 1

Citer

Elena Natterer, Roman Engelhardt, Sebastian Hörl, Klaus Bogenberger. Machine Learning Surrogates for Optimizing Transportation Policies with Agent-Based Models. 12th Triennial Symposium on Transportation Analysis (TRISTAN XII), Jun 2025, Okinawa, Japan. ⟨hal-04892998⟩

Collections

159 Consultations
0 Téléchargements

Partager

  • More