affiliation not provided to SSRN
Value decomposition, Multi-agent deep reinforcement learning, Robust regression, Maximum correntropy criterion