Dynamic Linear Bellman Combination of Optimal Policies for Solving New Tasks.
Takamitsu MatsubaraTakaya AsakuraKenji SugimotoPublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2015)
Keyphrases
- optimal policy
- markov decision processes
- semi markov decision processes
- decision problems
- dynamic programming algorithms
- dynamic programming
- finite horizon
- piecewise linear
- long run
- reinforcement learning
- state space
- finite state
- infinite horizon
- markov decision problems
- state dependent
- multistage
- policy iteration
- serial inventory systems
- markov decision process
- sufficient conditions
- average reward reinforcement learning
- evaluation function
- model checking
- initial state