Alexander L. Strehl
(Computer Science)

PAC Exploration in Reinforcement Learning

The subfield of artificial intelligence called reinforcement learning is expected to lead to smart technologies that allow devices to learn complex tasks on their own and adapt to the changing needs of customers. Mr. Strehl is credited with being the critical member of a team that has made a significant advance in creating efficient approaches to the exploration-exploitation dilemma—that is, how does a computer balance the need to learn from the environment against the need to accomplish its task.