Policy Expansion for Bridging Offline-to-Online Reinforcement Learning.
Haichao ZhangWe XuHaonan YuPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- optimal policy
- action selection
- policy search
- markov decision process
- real time
- exploration exploitation tradeoff
- online learning
- state space
- partially observable environments
- policy iteration
- policy gradient
- action space
- continuous state spaces
- continuous state
- markov decision problems
- partially observable markov decision processes
- control problems
- machine learning
- model free
- markov decision processes
- function approximation
- temporal difference
- learning algorithm
- policy evaluation
- decision problems
- control policies
- function approximators
- infinite horizon
- partially observable
- control policy
- batch mode
- state dependent
- optimal control
- finite state
- mobile robot
- e learning
- asymptotically optimal
- balancing exploration and exploitation