Reinforcement Learning for Mixing Loop Control with Flow Variable Eligibility Trace.
Anders OvergaardBrian Kongsgaard NielsenCarsten Skovmose KallesøeJan Dimon BendtsenPublished in: CCTA (2019)
Keyphrases
- reinforcement learning
- control problems
- optimal control
- control system
- function approximation
- robot control
- robotic systems
- control method
- feedback loop
- dynamic programming
- adaptive control
- hidden markov models
- database
- control strategies
- robotic control
- control policy
- flow patterns
- model free
- learning problems
- data acquisition
- state space
- learning process
- multi agent
- machine learning
- data mining