Control-Tutored Reinforcement Learning: an application to the Herding Problem

11/26/2019

∙

by   Francesco De Lellis, et al.

∙

0

∙

share

In this extended abstract we introduce a novel control-tutored Q-learning approach (CTQL) as part of the ongoing effort in developing model-based and safe RL for continuous state spaces. We validate our approach by applying it to a challenging multi-agent herding control problem.