Reinforcement Learning Meets Hybrid Zero Dynamics: A Case Study for RABBIT.
Guillermo A. CastilloBowen WengAyonga HereidZheng WangWei ZhangPublished in: ICRA (2019)
Keyphrases
- reinforcement learning
- case study
- state space
- dynamic model
- reinforcement learning algorithms
- test bed
- model free
- function approximation
- neural network
- hybrid approaches
- dynamical systems
- machine learning
- robot control
- collective behavior
- highly nonlinear
- multi agent reinforcement learning
- initial conditions
- learning capabilities
- human subjects
- temporal difference
- action selection
- hybrid learning
- learning problems
- markov decision processes
- multi agent
- bayesian networks
- real world