Learn Continuously, Act Discretely: Hybrid Action-Space Reinforcement Learning For Optimal Execution.
Feiyang PanTongzhe ZhangLing LuoJia HeShuoling LiuPublished in: IJCAI (2022)
Keyphrases
- action space
- reinforcement learning
- state space
- markov decision processes
- function approximators
- control policies
- continuous state spaces
- state and action spaces
- continuous state
- dynamic programming
- real valued
- function approximation
- action selection
- optimal control
- reinforcement learning methods
- learning algorithm
- stochastic processes
- learning agent
- optimal solution
- state information
- state action
- control policy
- reinforcement learning algorithms
- average cost
- model free
- heuristic search
- supervised learning
- search space
- multi agent
- average reward
- policy search
- bayesian networks