论文信息 - Estimating Gaussian Mixture Densities with EM

Estimating Gaussian Mixture Densities with EM – A Tutorial

Expectation Maximization (EM) [4, 3, 6] is a numerical algorithm for the maximization of functions of several variables. There are several tutorial introductions to EM, including [8, 5, 2, 7]. These are excellent references for greater generality about EM, several good intuitions, and useful explanations. The purpose of this document is to explain in a more self-contained way how EM can solve a special but important problem, the estimation of the parameters of a mixture of Gaussians from a set of data points. Here is the outline of what follows:

C. Tomasi

[1] D. Rubin,et al. Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[2] R. Okafor. Maximum likelihood estimation from incomplete data , 1987 .

[3] Heekuck Oh,et al. Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[4] Yair Weiss,et al. Motion Segmentation using EM - a short tutorial , 1996 .

[5] T. Minka. Expectation-Maximization as lower bound maximization , 1998 .

[6] Geoffrey E. Hinton,et al. A View of the Em Algorithm that Justifies Incremental, Sparse, and other Variants , 1998, Learning in Graphical Models.