EM for linear system identification

The aim of the code is to show how the EM algorithm can be used in the context of dynamic systems and hence, the example solved is very specific in order to illustrate the technique. 

MATLAB code

The MATLAB code is available here.

Relevant paper

All the details explaining the use of the expectation maximization (EM) algorithm for solving this particular example are provided in the follwing technical report,

Thomas B. Schön, An explanation of the Expectation Maximization algorithm. Division of Automatic Control, Linköping University, Sweden, Technical Report nr: LiTH-ISY-R-2915, August 2009. [pdf]

 © Thomas Schön 2022