Machine Learning Street Talk (MLST) cover image

#114 - Secrets of Deep Reinforcement Learning (Minqi Jiang)

Machine Learning Street Talk (MLST)

00:00

Understanding Learning Potential in Reinforcement Learning

This chapter explores the concepts of regret and value prediction errors in assessing learning potential for agents in reinforcement learning. It highlights a recent study in a 3D environment, discusses various regret estimators, and examines the challenges in selecting optimal training criteria.

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app