Risk-Sensitive Reinforcement Learning: a Martingale Approach to Reward Uncertainty.
Nelson VadoriSumitra GaneshPrashant P. ReddyManuela VelosoPublished in: CoRR (2020)
Keyphrases
- risk sensitive
- reinforcement learning
- model free
- optimal control
- markov decision processes
- markov decision problems
- control policies
- expected utility
- function approximation
- reinforcement learning algorithms
- average reward
- utility function
- markov decision chains
- reward function
- temporal difference
- state space
- optimal policy
- dynamic programming
- learning algorithm
- policy iteration
- average cost
- machine learning
- multi agent
- action space
- control policy
- control strategies
- belief functions
- infinite horizon
- finite state
- sufficient conditions
- supervised learning
- belief revision
- action selection
- finite horizon
- function approximators
- decision theory