Multiresolution state-space discretization method for Q-Learning.
Amanda Kathryn LamptonJohn ValasekPublished in: ACC (2009)
Keyphrases
- state space
- discretization method
- multiresolution
- discretization methods
- continuous attributes
- reinforcement learning
- reinforcement learning algorithms
- discrete data
- preprocessing step
- optimal policy
- naive bayes classifier
- state variables
- dynamic programming
- markov decision processes
- search space
- dynamical systems
- continuous state spaces
- particle filter
- model free
- continuous data
- data mining
- data sets
- action space
- data streams