Risk-averse dynamic programming for Markov decision processes.
Andrzej RuszczynskiPublished in: Math. Program. (2010)
Keyphrases
- markov decision processes
- risk averse
- dynamic programming
- stochastic programming
- multistage
- optimal policy
- utility function
- state space
- decision makers
- finite state
- linear program
- reinforcement learning
- portfolio management
- inventory level
- average cost
- finite horizon
- optimal control
- infinite horizon
- linear programming
- policy iteration
- action space
- expected utility
- partially observable
- decision processes
- risk sensitive
- markov decision process
- reward function
- average reward
- decision theory
- decision problems
- search algorithm