University of Massachusetts Lowell Multi-agent Multi-Armed Bandits, Multi-Agent Reinforcement Learning, Learning from Demonstration, Robot Learning