Hangzhou

The 'Hangzhou' dataset is a benchmark that contains large-scale traffic signal control data from 16 intersections, used to evaluate the performance of Cooperative Multi-Agent Reinforcement Learning (MARL) approaches.

Papers using Hangzhou (3)

DGLight: DQN-Guided GRPO Fine-Tuning of Large Language Models for Traffic Signal Control2026MAESTRO: Multi-Agent Environment Shaping through Task and Reward Optimization2025MoveLight: Enhancing Traffic Signal Control through Movement-Centric Deep Reinforcement Learning2024