Markov decision problem based on a two-level system
by Huang Jin
THESIS
2003
M.Phil. Electrical and Electronic Engineering
vi, 35 leaves ; 30 cm
Abstract
Many problems in discrete event dynamic systems(DEDS) can be modeled as Markov Decision Processes. However the computation of a MDP problem increase rapidly with the increase of the state space. In this thesis, we consider the markov decision problem based on a two-level system. Although the two-level system may have many states in its state space, we can give an efficient policy iteration algorithm to find the optimal policy.
Post a Comment