Hierarchical Policies of Subgoals for Safe Deep Reinforcement Learning.
Fumin YuFeng GaoYao YuanXiaofei XingYinglong DaiPublished in: UbiSec (2022)
Keyphrases
- reinforcement learning
- hierarchical reinforcement learning
- state abstraction
- optimal policy
- policy search
- reinforcement learning agents
- state space
- reward function
- control policies
- semi markov decision process
- fitted q iteration
- markov decision process
- markov decision processes
- machine learning
- function approximation
- model free
- policy gradient methods
- learning algorithm
- markov decision problems
- total reward
- robotic control
- reinforcement learning algorithms
- multiple layers
- transfer learning
- deep learning
- decentralized control
- partially observable markov decision processes
- infinite horizon
- coarse to fine
- hierarchical clustering
- sufficient conditions
- dynamic programming
- genetic algorithm