Multi-agent reinforcement learning using echo-state network and its application to pedestrian dynamics

arXiv:2312.11834v2 Announce Type: replace-cross
Abstract: In recent years, simulations of pedestrians using the multi-agent reinforcement learning (MARL) have been studied. This study considered the roads on a grid-world environment, and implemented pedestrians as MARL agents using an echo-state network and the least squares policy iteration method. Under this environment, the ability of these agents to learn to move forward by avoiding other agents was investigated. Specifically, we considered two types of tasks: the choice between a narrow direct route and …

abstract agent agents application arxiv cs.ai cs.lg cs.ma dynamics echo environment grid iteration least multi-agent network pedestrian pedestrians physics.soc-ph policy reinforcement reinforcement learning roads simulations squares state study type world