The objective of reinforcement learning is to discover a policy, that is a mapping from states to steps, that maximizes the anticipated cumulative reward over time. A collaboration across continents to resolve a plastics issue MIT pupils travel into the Amazon, working with locals to deal with th