First Order Optimization in Policy Space for Constrained Deep Reinforcement Learning.
Yiming ZhangQuan VuongKeith W. RossPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- optimal policy
- action space
- optimization algorithm
- first order logic
- function approximation
- action selection
- space time
- global optimization
- concave convex procedure
- policy search
- markov decision process
- constrained optimization
- optimization process
- markov decision processes
- function approximators
- control policy
- state and action spaces
- temporal difference
- partially observable
- optimization method
- control policies
- policy gradient
- policy evaluation
- optimization problems
- search space