Reinforcement learning for Golog programs with first-order state-abstraction.
Daniel BeckGerhard LakemeyerPublished in: Log. J. IGPL (2012)
Keyphrases
- state abstraction
- reinforcement learning
- state space
- agent programming
- reinforcement learning agents
- initial state
- situation calculus
- markov decision processes
- reinforcement learning algorithms
- path finding
- action theories
- hierarchical reinforcement learning
- partial observability
- first order logic
- function approximation
- decision theoretic
- multi agent
- machine learning
- temporal difference
- model free
- optimal policy
- supervised learning
- function approximators
- markov decision process
- learning algorithm