Two Decentralised Learning Problems - Sketching and Policy Evaluation - Justin Romberg, Georgia Tech

Explore two decentralized learning problems in this 46-minute workshop talk by Justin Romberg from Georgia Tech. Delve into the topics of sketching and policy evaluation as part of the Isaac Newton Institute's program on "Approximation, sampling and compression in data science." Gain insights from leading researchers in mathematics, statistics, computer science, and engineering as they discuss cutting-edge advances in data science. Discover how this workshop aims to bridge gaps between various communities working on mathematical aspects of data science, including computational statistics, machine learning, optimization, information theory, and learning theory. Engage with a forum designed to stimulate collaboration and exchange ideas among researchers in these interconnected fields.

Two decentralised learning problems: Sketching and policy evaluation - Justin Romberg, Georgia Tech