Matrix-Valued Mean-Field-Type Games: Risk-Sensitive, Adversarial, and Risk-Neutral Linear-Quadratic Case.
Julian Barreiro-GomezTyrone E. DuncanHamidou TembinePublished in: CoRR (2019)
Keyphrases
- risk sensitive
- risk neutral
- optimal control
- utility function
- risk averse
- dynamic programming
- vector valued
- markov decision processes
- optimality criterion
- decision theoretic
- expected utility
- model free
- em algorithm
- risk aversion
- markov random field
- control policies
- reinforcement learning
- game theory
- average cost
- efficient optimization