An approximate dynamic programming approach to probabilistic reachability for stochastic hybrid systems.
Alessandro AbateMaria PrandiniJohn LygerosShankar SastryPublished in: CDC (2008)
Keyphrases
- approximate dynamic programming
- hybrid systems
- stochastic dynamic programming
- discrete event
- linear program
- reinforcement learning
- soft computing
- intelligent systems
- dynamic programming
- meta learning
- neural network
- computational intelligence
- black box
- artificial intelligence
- state space
- control policy
- step size
- machine learning
- dynamic systems
- bayesian networks
- least squares
- average cost
- simulation model
- policy iteration
- artificial neural networks
- inductive learning
- semi supervised