Distributional Reinforcement Learning for Risk-Sensitive Policies.
Shiau Hong LimIlyas MalikPublished in: NeurIPS (2022)
Keyphrases
- risk sensitive
- reinforcement learning
- control policies
- optimal policy
- markov decision problems
- markov decision processes
- model free
- optimal control
- control policy
- state space
- average cost
- action space
- markov decision process
- function approximation
- partially observable
- reinforcement learning algorithms
- reward function
- finite horizon
- policy iteration
- optimality criterion
- multi agent
- average reward
- partially observable markov decision processes
- control strategies
- dynamic programming
- markov decision chains
- temporal difference
- infinite horizon
- long run
- finite state
- linear programming
- learning algorithm
- machine learning
- learning capabilities
- decision theoretic
- decision processes
- decision problems
- supervised learning
- search space