ADSI Summer Workshop- Algorithmic Foundations of Learning and Control
Paul G. Allen School via YouTube
Overview
Explore the cutting-edge research on stable and efficient reinforcement learning algorithms in this 54-minute conference talk presented by Csaba Szepesvári from the University of Alberta at the 2019 ADSI Summer Workshop. Delve into the topic of "Politex - Towards Stable and Efficient Reinforcement Learning Algorithms that Generalize" as Szepesvári discusses key concepts including notation, pseudo regret, measureless difference, online learning, and linear prediction. Gain insights into the latest results and engage in a thought-provoking discussion on the future of reinforcement learning algorithms that can generalize effectively.
Syllabus
Introduction
Notation
Pseudo regret
Measureless difference
Online learning
Linear prediction
Results
Discussion
Taught by
Paul G. Allen School