World Simulation
Reinforcement learning
Just like humans, classical reinforcement learning struggles to identify the connections between critical events and drastic consequences that only manifest much later. We investigate adaptive approaches to augment the classic algorithm.