About this Event
The Expectation Maximization algorithm (EM) figures prominently in current neural networks and neuroscience work. We shall present the basics of EM in a maximum likelihood framework as in Dempster, Laird, Rubin (1977). We shall also examine convergence properties of the algorithm as in Wu (1982), and the more recent approach via free energy of Neal and Hinton (1998).
0 people are interested in this event