ConceptAI & ML
Partially observable Markov decision process
A POMDP is a generalization of a Markov decision process that adds unobserved states, observations, and an observation function where the agent maintains a belief state distribution over states.
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.
Lecturers
3Each in their own terms, from the easiest level.
David Silver
The formalism used when working with partially observable environments where the agent must construct its own state representation.Sergey Levine
A POMDP is an MDP that has observations and observation probabilities in addition to states and actions.
Connections
8The lectures that define it or assume it, and the concepts next to it.