Discrete-time decentralized control using the risk-sensitive performance criterion in the large population regime: A mean field approach.
Jun MoonTamer BasarPublished in: ACC (2015)
Keyphrases
- risk sensitive
- decentralized control
- markov decision processes
- finite state
- optimality criterion
- state space
- optimal policy
- reinforcement learning
- multiagent systems
- optimal control
- utility function
- average reward
- smart grid
- infinite horizon
- average cost
- decision makers
- markov chain
- reinforcement learning algorithms
- goal oriented
- dynamic programming
- virtual communities
- markov decision problems
- expected utility
- genetic algorithm
- reward function
- model free
- belief networks
- evaluation function
- decision problems