Policy Learning for Nonlinear Model Predictive Control With Application to USVs.

QFree: A Universal Value Function Factorization for Multi-Agent Reinforcement Learning.

Attention Loss Adjusted Prioritized Experience Replay.