Here is a illustration of Expectation maximization Algorithm for Gaussian mixtures from C. M. Bishop's book.

Mixture OldKiwi.jpg

As can be seen from the figures, algorithm, starts with randomly initialized parameters. At each iteration, based on the current parameters, it clusters the data. Using that clustered data, the algorithm computes the parameters again. The process continues until convergence.

Alumni Liaison

To all math majors: "Mathematics is a wonderfully rich subject."

Dr. Paul Garrett