Multi-resolution state-space discretization for Q-Learning with pseudo-randomized discretization.
Amanda Kathryn LamptonJohn ValasekMrinal KumarPublished in: IJCNN (2010)
Keyphrases
- state space
- multiresolution
- reinforcement learning
- heuristic search
- markov decision processes
- iterative refinement
- optimal policy
- reinforcement learning algorithms
- dynamic programming
- preprocessing
- discretization methods
- discretization method
- state variables
- neural network
- function approximation
- learning rate
- dynamical systems
- continuous variables
- naive bayes classifier
- multi agent
- discrete space
- cooperative
- search space
- planning problems
- initial state
- wavelet transform