Using Deep Reinforcement Learning to Learn High-Level Policies on the ATRIAS Biped.
Tianyu LiHartmut GeyerChristopher G. AtkesonAkshara RaiPublished in: ICRA (2019)
Keyphrases
- reinforcement learning
- high level
- optimal policy
- reinforcement learning agents
- hierarchical reinforcement learning
- low level
- policy search
- markov decision process
- markov decision processes
- state space
- function approximators
- tabula rasa
- control policies
- programming language
- learning agent
- reinforcement learning algorithms
- function approximation
- optimal control
- model free
- control strategy
- higher level
- learning algorithm
- multi agent
- reward function
- lower level
- low level features
- policy gradient methods
- control system
- relevance feedback
- semantic information
- decision problems
- partially observable markov decision processes
- temporal difference
- action selection
- machine learning
- control law