Value Function Discovery in Markov Decision Processes With Evolutionary Algorithms.
Martijn OnderwaterSandjai BhulaiRob van der MeiPublished in: IEEE Trans. Syst. Man Cybern. Syst. (2016)
Keyphrases
- markov decision processes
- evolutionary algorithm
- finite state
- optimal policy
- state space
- reinforcement learning
- policy iteration
- reachability analysis
- dynamic programming
- transition matrices
- action space
- multi objective
- average reward
- infinite horizon
- decision theoretic planning
- finite horizon
- planning under uncertainty
- markov decision process
- reinforcement learning algorithms
- multi objective optimization
- factored mdps
- average cost
- risk sensitive
- model based reinforcement learning
- partially observable
- decision processes
- neural network
- differential evolution
- state and action spaces
- multi agent
- genetic algorithm
- data mining