<ul><li>Reinforcement learning methods have proposed promising traffic signal control policy that can be trained on large road networks.</li><li>MacLight, Multi-Scene Aggregation Convolutional Learning, offers faster training speeds and more stable performance.</li><li>It utilizes variational autoencoders for global representation and proximal policy optimization algorithm for value evaluation.</li><li>Experimental results demonstrate superior stability, optimized convergence levels, and the highest time efficiency.</li></ul>

MacLight: Multi-scene Aggregation Convolutional Learning for Traffic Signal Control

Discover more