Opening Foundational concepts in deep reinforcement learning
Foundational concepts in deep reinforcement learning
- Length
- 19 min
- Level
- Undergraduate
- Recorded
- UC Berkeley2020
Chapter 1 of 4: Before the lecture
Model-based reinforcement learning starts from transition dynamics to make decisions when rules or physical laws are known. The lecture distinguishes between open-loop planning, which commits to a fixed sequence of actions, and closed-loop planning, which conditions actions on observed states. While open-loop planning is straightforward in deterministic settings, it fails in stochastic environments where future information changes the optimal response. The material introduces optimal control, trajectory optimization, and system identification, setting up methods that use known models before later extending them to learned models.

Your comment is kept while you sign up.Sign in
Contrasts open-loop and closed-loop planning to show why open-loop action sequences fail in stochastic environments.
Model-based reinforcement learning starts from transition dynamics to make decisions when rules or physical laws are known. The lecture distinguishes between open-loop planning, which commits to a fixed sequence of actions, and closed-loop planning, which conditions actions on observed states. While open-loop planning is straightforward in deterministic settings, it fails in stochastic environments where future information changes the optimal response. The material introduces optimal control, trajectory optimization, and system identification, setting up methods that use known models before later extending them to learned models.
This lecture
Lecture 28 of 73 of Deep Reinforcement Learning.
Choose a node to see how it connects to this lecture. What comes before is on the leftat the top. What comes after is on the rightat the bottom. The ring is what it teaches. People and places are above. Other lectures and books are below.
Press a chosen node again to open it.