Risk-Sensitive Policy with Distributional Reinforcement Learning.
Thibaut ThéateDamien ErnstPublished in: Algorithms (2023)
Keyphrases
- risk sensitive
- reinforcement learning
- control policies
- markov decision problems
- optimal policy
- markov decision processes
- model free
- optimal control
- policy iteration
- control policy
- action space
- average cost
- state space
- markov decision process
- partially observable
- function approximation
- average reward
- reinforcement learning algorithms
- infinite horizon
- finite horizon
- reward function
- decision processes
- optimality criterion
- temporal difference
- dynamic programming
- finite state
- action selection
- utility function
- long run
- multistage
- function approximators
- supervised learning
- learning algorithm
- machine learning
- decision problems
- linear programming
- bayesian networks
- partially observable markov decision processes
- decision theoretic
- control strategy
- real valued
- np hard
- multi agent