An adaptive state aggregation approach to Q-learning with real-valued action function.
Kao-Shing HwangYu-Jen ChenPublished in: SMC (2010)
Keyphrases
- real valued
- action space
- state action
- state space
- function approximators
- reinforcement learning
- complex valued
- interval valued
- continuous functions
- reproducing kernel hilbert space
- distributed representations
- action selection
- integer valued
- multiple instance
- neural network
- search engine
- function approximation
- ranking functions
- co occurrence
- supervised learning
- prior knowledge
- digital libraries
- agent learns
- learning algorithm