- Article
It is challenging to ensure satisfying co-evolution efficiency for the multi-agents in dynamic environments since during Actor-Critic training there is a high probability of falling into local optimality, failing to adapt to the suddenly changed envi...