Necessary and sufficient conditions for a solution to the risk-sensitive Poisson equation on a finite state space.
Rolando Cavazos-CadenaDaniel Hernández-HernándezPublished in: Syst. Control. Lett. (2009)
Keyphrases
- sufficient conditions
- finite state
- risk sensitive
- markov decision processes
- optimal policy
- boundary conditions
- poisson equation
- markov decision chains
- average cost
- linear systems
- markov chain
- boundary value problem
- model checking
- stationary policies
- markov decision problems
- fixed point
- action space
- state space
- mathematical model
- policy iteration
- optimal control
- gradient domain
- infinite horizon
- control policies
- markov decision process
- partially observable markov decision processes
- long run
- exact solution
- reinforcement learning