Learning from user's behaviour of some well-known congested traffic networks

TL;DR

This paper explores using machine learning, especially Graph Neural Networks, to efficiently predict traffic flow distributions in congested networks, reducing computational costs while maintaining accuracy.

Contribution

It introduces a novel ML-based approach leveraging GNNs to approximate traffic assignment solutions more efficiently than traditional iterative methods.

Findings

GNN-based models achieve comparable accuracy to traditional methods.

The approach significantly reduces computation time for traffic prediction.

The method demonstrates robustness across different traffic network scenarios.

Abstract

The traffic assignment problem (TAP) aims to predict how traffic flows distribute themselves across a road network, traditionally requiring computationally expensive iterative simulations to reach a user equilibrium (UE) where no driver can unilaterally reduce their travel time. Recent developments in machine learning (ML), particularly Graph Neural Networks (GNNs) and hybrid approaches, aim to solve this faster while maintaining accuracy