Continuous Control With a Combination of Supervised and Reinforcement Learning.
Dmitry KanginNicolas PugeaultPublished in: IJCNN (2018)
Keyphrases
- continuous state spaces
- control problems
- reinforcement learning
- optimal control
- action space
- markov decision processes
- continuous state
- learning algorithm
- adaptive control
- supervised learning
- multi agent
- machine learning
- action selection
- control strategies
- function approximation
- control strategy
- optimal policy
- control system
- feature selection
- real time
- combining multiple
- model free
- robot control
- control policy
- control method
- data acquisition
- unsupervised learning
- control theory
- dynamic programming
- function approximators