A Hybrid Genetic/Optimization Algorithm for Finite-Horizon, Partially Observed Markov Decision Processes.
Zong-Zhi LinJames C. BeanChelsea C. White IIIPublished in: INFORMS J. Comput. (2004)
Keyphrases
- optimization algorithm
- finite horizon
- partially observed
- markov decision processes
- hybrid genetic
- optimal policy
- multi objective
- optimal stopping
- infinite horizon
- average cost
- finite state
- state space
- reinforcement learning
- markov decision process
- policy iteration
- differential evolution
- transition matrices
- dynamic programming
- partially observable
- action space
- control policies
- modeling method
- long run
- expected reward
- support vector regression
- cost function
- evolutionary algorithm
- decision making
- learning algorithm
- machine learning
- search algorithm
- support vector
- genetic algorithm