Video
Description
Neural Q-Learning builds on the theory developed in previous sessions, augmenting the tabular Q-Learning algorithm with the powerful function approximation capabilities of Neural Networks. NQL is the “base” algorithm unifying Neural Networks and Reinforcement Learning, and participants will be exposed to both the impressive generalization properties of this algorithm, as well as some of it’s potential drawbacks and limitations.
Lecture Slides
StarAi Lecture 5 part 1 Neural Q Theory slides
StarAi Lecture 5 part 2 Neural Q Implementation slides
Exercise
Follow the link below to access the exercises for lecture 5:
lecture 5 Exercise: Neural Q Learning Exercise
Exercise Solutions
Follow the link below to access the exercise solutions for lecture 5:
lecture 5 Exercise: Neural Q Learning Exercise Solutions
Additional Learning Material
- Sutton & Barto’s Reinforcement Learning: An Introduction - Chapter 16 section 16.5