Count-based Exploration in Reinforcement Learning
GPTKB entity
Statements (23)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:exploration_method
|
| gptkbp:alternativeTo |
entropy-based exploration
parameter noise exploration |
| gptkbp:application |
deep RL
tabular RL |
| gptkbp:approach |
assigns bonus based on state visitation counts
|
| gptkbp:challenge |
scaling to high-dimensional state spaces
|
| gptkbp:citation |
Bellemare et al. 2016
Strehl and Littman 2008 |
| gptkbp:field |
gptkb:reinforcement_learning
|
| gptkbp:goal |
encourage exploration of less-visited states
|
| gptkbp:influencedBy |
gptkb:multi-armed_bandit_problem
|
| gptkbp:notableFor |
gptkb:E3
MBIE-EB Pseudo-counts R-MAX |
| gptkbp:relatedTo |
exploration-exploitation tradeoff
intrinsic motivation |
| gptkbp:solvedBy |
density models for pseudo-counts
|
| gptkbp:usedIn |
gptkb:machine_learning
|
| gptkbp:bfsParent |
gptkb:Hari_Vedantam
|
| gptkbp:bfsLayer |
8
|
| https://www.w3.org/2000/01/rdf-schema#label |
Count-based Exploration in Reinforcement Learning
|