On the global convergence of relative value iteration for infinite-horizon risk-sensitive control of diffusions.
Hassan HmediAri ArapostathisGuodong PangPublished in: Syst. Control. Lett. (2023)
Keyphrases
- infinite horizon
- optimal control
- risk sensitive
- global convergence
- markov decision chains
- markov decision processes
- control strategy
- dynamic programming
- average cost
- finite horizon
- optimization methods
- convergence rate
- global optimum
- partially observable
- convergence speed
- markov decision process
- optimal policy
- reinforcement learning
- control system
- long run
- mathematical model
- policy iteration
- cost function
- control policies
- optimal solution
- markov decision problems
- control policy
- state space
- control strategies
- multistage