Kinematic State Abstraction and Provably Efficient Rich-Observation Reinforcement Learning

KinematicStateAbstractionandProvablyEfficientRich-ObservationReinforcementLearningDipendraMisra1MikaelHenaff1AkshayKrishnamurthy1JohnLangford1Abstractfromthewell-studiedtabularsettingtoexploretheen...