Monte Carlo TD(λ)-methods for the optimal control of discrete-time Markovian jump linear systems.
Oswaldo L. V. CostaJulio C. C. AyaPublished in: CDC (2000)
Keyphrases
- monte carlo
- linear systems
- optimal control
- markov chain
- temporal difference
- optimal control problems
- control theory
- dynamical systems
- control problems
- finite state
- state space
- dynamic programming
- reinforcement learning
- sufficient conditions
- control strategy
- function approximation
- reinforcement learning algorithms
- policy evaluation
- importance sampling
- infinite horizon
- control law
- policy iteration
- particle filter
- mathematical model
- function approximators
- reinforcement learning methods
- real time
- machine learning
- control method
- support vector
- artificial neural networks
- linear programming