Reinforcement Learning through Global Stochastic Search in N-MDPs.
Matteo LeonettiLuca IocchiSubramanian RamamoorthyPublished in: ECML/PKDD (2) (2011)
Keyphrases
- stochastic search
- reinforcement learning
- markov decision processes
- state space
- simulated annealing
- optimal policy
- variable selection
- function approximation
- consistency checking
- search methods
- partially observable
- continuous state and action spaces
- machine learning
- reinforcement learning algorithms
- temporal difference
- policy iteration
- multi agent
- policy search
- model free
- markov decision process
- genetic algorithm
- heuristic search
- dynamic programming
- evolutionary algorithm
- tree search
- learning algorithm