partially observable Markov decision process
GPTKB entity
Statements (50)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:logic
|
| gptkbp:abbreviation |
gptkb:POMDP
|
| gptkbp:field |
gptkb:artificial_intelligence
gptkb:reinforcement_learning decision theory operations research |
| gptkbp:form |
tuple (S, A, O, T, Z, R, γ)
|
| gptkbp:generalizes |
Markov chain
|
| gptkbp:hasApplication |
gptkb:dialogue_systems
gptkb:game_AI autonomous vehicles resource management diagnosis |
| gptkbp:hasComponent |
states
actions reward function observation model observations transition model |
| gptkbp:hasProperty |
hidden state
non-deterministic outcomes partial observability sequential decision making stochastic observations stochastic transitions |
| https://www.w3.org/2000/01/rdf-schema#label |
partially observable Markov decision process
|
| gptkbp:introduced |
gptkb:Ronald_A._Howard
|
| gptkbp:introducedIn |
1960s
|
| gptkbp:limitation |
gptkb:curse_of_dimensionality
computational complexity curse of history |
| gptkbp:parameter |
gptkb:award
gptkb:public_policy discount factor belief state initial belief observation probability transition probability |
| gptkbp:relatedTo |
Markov chain
belief state |
| gptkbp:solvedBy |
gptkb:Monte_Carlo_methods
policy iteration value iteration point-based value iteration |
| gptkbp:usedFor |
planning
robotics control modeling decision making under uncertainty |
| gptkbp:bfsParent |
gptkb:Markov_chain
|
| gptkbp:bfsLayer |
5
|