Sample-Efficient Exploration in Reinforcement Learning with Rich Observations - 2019

Sample-Efficient Exploration in Reinforcement Learning with Rich Observations - 2019

Explore sample-efficient exploration techniques in reinforcement learning with rich observations in this 46-minute conference talk presented by Alekh Agarwal from Microsoft Research. Delivered at the 2019 ADSI Summer Workshop on Algorithmic Foundations of Learning and Control, hosted by the Paul G. Allen School of Computer Science & Engineering at the University of Washington, the talk delves into advanced strategies for improving learning efficiency in complex reinforcement learning environments. Gain insights into cutting-edge research that addresses the challenges of learning from high-dimensional observations while maintaining sample efficiency, a crucial aspect for practical applications of reinforcement learning algorithms.