New prioritized value iteration for Markov decision processes.
Ma. de Guadalupe García-HernándezJosé Ruiz-PinalesEva OnaindiaJuan Gabriel Aviña-CervantesSergio E. Ledesma-OrozcoEdgar Alvarado-MendezAlberto Reyes-BallesterosPublished in: Artif. Intell. Rev. (2012)
Keyphrases
- markov decision processes
- policy iteration
- finite state
- state space
- reinforcement learning
- dynamic programming
- optimal policy
- average reward
- transition matrices
- decision theoretic planning
- model based reinforcement learning
- infinite horizon
- reinforcement learning algorithms
- state and action spaces
- decision processes
- factored mdps
- markov decision process
- action space
- risk sensitive
- average cost
- planning under uncertainty
- reward function
- action sets
- partially observable
- state abstraction
- decision diagrams
- multistage
- machine learning
- stochastic shortest path