Learning Algorithms for Markov Decision Processes with Average Cost.
Jinane AbounadiDimitri P. BertsekasVivek S. BorkarPublished in: SIAM J. Control. Optim. (2001)
Keyphrases
- average cost
- markov decision processes
- learning algorithm
- reinforcement learning algorithms
- reinforcement learning
- finite state
- optimal policy
- state space
- approximate dynamic programming
- dynamic programming
- finite horizon
- infinite horizon
- transition matrices
- average reward
- markov decision chains
- partially observable
- decision theoretic planning
- risk sensitive
- policy iteration
- machine learning
- markov decision process
- planning under uncertainty
- action space
- long run
- learning tasks
- control policy
- action sets
- markov decision problems
- initial state
- learning rate
- learning agent
- optimal control
- linear program
- reward function
- stationary policies
- finite number
- supervised learning