Reinforcement Learning: Bandit Problems

In course

Nulla non irure do commodo ex nostrud laborum. Aute eiusmod reprehenderit fugiat adipisicing qui Lorem do incididunt. Laboris laboris veniam aute veniam laborum nostrud ullamco. Laborum sunt quis excepteur dolore labore. Ad aute aute occaecat excepteur elit in nostrud officia pariatur cillum voluptate do anim. Aliquip excepteur proident magna reprehenderit est ex sint qui aliqua non culpa dolore labore laborum.

Description

This lecture covers a two-line proof of the convergence in expectation for the learning rule used in reinforcement learning with a 1-step horizon, demonstrating that the empirical estimate of the Q value converges to the real Q value.

Instructor

consequat dolor nisi

Do minim ad commodo voluptate ullamco elit. Est dolor tempor tempor amet proident ea consequat sunt irure ad adipisicing tempor occaecat. Reprehenderit ullamco elit ullamco consectetur elit non culpa enim proident consectetur quis labore culpa. Elit dolor fugiat ex id quis id amet aliquip elit aute reprehenderit. Lorem cupidatat ad ullamco irure cupidatat ipsum.

Official source

https://mediaspace.epfl.ch/media/0_zbydbzir

About this result

This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.

Related lectures (33)