Intervention Aided Reinforcement Learning for Safe and Practical Policy Optimization in Navigation.
Fan WangBo ZhouKe ChenTingxiang FanXi ZhangJiangyong LiHao TianJia PanPublished in: CoRR (2018)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- action selection
- markov decision process
- partially observable
- approximate dynamic programming
- actor critic
- optimization algorithm
- markov decision processes
- control problems
- reward function
- optimization process
- model free
- control policy
- policy iteration
- action space
- function approximators
- infinite horizon
- control policies
- function approximation
- global optimization
- optimization method
- dynamical systems
- optimization problems
- dynamic programming
- policy gradient
- average reward
- real world
- partially observable environments
- reinforcement learning algorithms
- long run
- indoor environments
- constrained optimization
- state space
- multi agent
- learning algorithm