Explore offline deep reinforcement learning algorithms in this 32-minute lecture by Sergey Levine from UC Berkeley. Delve into the workings of modern machine learning, examining concepts like overfitting, distributional shift, and implicit constraints. Learn about conservative Q-learning and the D4RL dataset. Gain insights into the latest results and conclusions in this field, enhancing your understanding of deep reinforcement learning techniques and their applications.
Offline Deep Reinforcement Learning Algorithms
Overview
Syllabus
Intro
Why does modern machine learning work
Overview
Overfitting
Distributional Shift
implicit constraints
conservative qlearning
d4rl
Results
Conclusion
Taught by
Simons Institute
Related Courses
-
Reinforcement Learning Course - Full Machine Learning Tutorial
-
Deep Reinforcement Learning in Python
-
Understanding Algorithms for Reinforcement Learning
-
AI skills: Introduction to Unsupervised, Deep and Reinforcement Learning
-
Fundamentals of Deep Reinforcement Learning
-
Deep Reinforcement Learning: Hands-on AI Tutorial in Python
