Output Feedback Q-Learning for Linear-Quadratic Discrete-Time Finite-Horizon Control Problems.
Giuseppe Carlo CalafioreCorrado PossieriPublished in: IEEE Trans. Neural Networks Learn. Syst. (2021)
Keyphrases
- optimal control
- linear quadratic
- control problems
- finite horizon
- infinite horizon
- reinforcement learning
- optimal policy
- dynamic programming
- average cost
- continuous state spaces
- markov decision processes
- state space
- control strategy
- markov decision process
- control policies
- learning algorithm
- control law
- finite state
- brownian motion
- multi agent
- multistage
- long run
- state dependent
- sufficient conditions