Full Gradient Deep Reinforcement Learning for Average-Reward Criterion.
Tejas PagareVivek S. BorkarKonstantin AvrachenkovPublished in: CoRR (2023)
Keyphrases
- average reward
- reinforcement learning
- optimality criterion
- policy gradient
- markov decision processes
- optimal policy
- model free
- actor critic
- semi markov decision processes
- stochastic games
- policy iteration
- long run
- state and action spaces
- state action
- function approximation
- total reward
- state space
- reinforcement learning algorithms
- hierarchical reinforcement learning
- rl algorithms
- dynamic programming
- discounted reward
- markov chain
- learning algorithm
- decision problems
- reinforcement learning methods
- multi agent
- partially observable markov decision processes
- reward function
- temporal difference
- optimal control
- transfer learning
- function approximators
- action selection
- reward shaping
- finite state