ConceptAI & ML
Temporal difference learning
A learning method that updates value estimates incrementally based on each observed transition rather than waiting for the end of an episode.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Opens up
Concepts that name this one as a prerequisite.
Lecturers
3Each in their own terms, from the easiest level.
David Silver
Methods that learn directly from experience in a model-free manner by updating guesses towards other guesses using incomplete episodes.Russ Tedrake
A reinforcement learning method that learns from bootstrapping off the current estimate of the value function.
Lectures that assume it
6Connections
18The lectures that define it or assume it, and the concepts next to it.
Concepts9
- Markov decision processes
- Bellman equation
- Policy evaluation
- Reinforcement learning
- Bootstrapping
- Action-value actor-critic