Reinforcement Learning: Frontier of Artificial Intelligence

Reinforcement learning (RL) is a sub-area of machine learning to solve control tasks such as self-driving, learning to playing Chess or Go. RL algorithms need to balance exploration (finding new ways) as well as  exploitation (sticking to old ways that have worked).

 

The first part of the talk will be generic in nature: decision theoretic foundations of artificial intelligence and machine learning will be presented, and the reasons behind the current excitement in deep learning will be discussed.

 

The second part of the talk will be about an on-going work on an RL algorithm for optimal curriculum progression.