Constrained Risk-Averse Markov Decision Processes.

Mohamadreza Ahmadi Ugo Rosolia Michel D. Ingham Richard M. Murray Aaron D. Ames

Published in: AAAI (2021)

Keyphrases

markov decision processes
risk averse
risk neutral
risk sensitive
decision makers
utility function
optimal policy
stochastic programming

finite state
reinforcement learning
dynamic programming
policy iteration
state space
portfolio management
transition matrices
markov decision process

expected utility
infinite horizon
partially observable
finite horizon
decision processes
average reward
average cost

partially observable markov decision processes
action space
reward function
linear program
multistage
optimal solution
learning algorithm