Improved Reinforcement Learning in Cooperative Multi-agent Environments Using Knowledge Transfer

no code implementations • 20 Jul 2021 • Mahnoosh Mahdavimoghaddam, Amin Nikanjam, Monireh Abdoos

In the proposed communication framework, agents learn to cooperate effectively and also by introduction of a new state calculation method the size of state space will decline considerably.

Multi-agent Reinforcement Learning reinforcement-learning +2