Workshop Overview

This workshop covers an overview of foundational theory and practical skills to train and evaluation a robot system using deep reinforcement learning.

Agenda

Core Topics

• Theoretical RL Presentation
  • Reinforcement Learning Intro - Basics of model free reinforcement learning, including Value and Q-functions, which are the basis for one class of algorithms classified as Actor-Critic algorithms.
  • Feature Based RL - How do we move from tabular to feature-based Actor-Critic algorithms.
  • Deep RL - Basics on DRL Actor-Critic Algorithms like the Soft-Actor-Critic algorithm (SAC).
• Coding
  • Stable Baselines - learning to train and evaluate. See our documentation.

Workshop Notebooks

Stable Baselines 3 (SB3)

Metrics and Logging

Workshop Slide Decks & Video

Slide Decks

https://drive.google.com/file/d/187GGig4YbXJ-8pSvgdnL8Z03r2f0bg86/view

Zoom Recording/Link

https://lipscomb.zoom.us/j/88478002720?pwd=wpK5FgcFMThFV5iXCekX7dxyG7O1s3.1

Password**:** 000000

Zoom Recording here. Passcode: y7T%rE^p

Follow Up