UKC

Statistics: max likelihood for mixture of gaussians

New Topic
This topic has been archived, and won't accept reply postings.
 Oujmik 28 Mar 2014
I've been reading Dave MacKay's excellent book on Information Theory - available here:

http://www.inference.phy.cam.ac.uk/mackay/itila/book.html

Problem is my maths skills are very rusty. I understand the concepts and find them fascinating but actually doing most of the exercises is beyond me. My main interest is in inference and I particularly want to solve problem 22.15 (page 309) but I can't immediately see how to solve the mixture of gaussians problem for distributions with a common mean and varying SD. I expect it would help if I'd been able to follow the derivation in section 22.3 for the iterative method with constant SD and varying means.

Can anyone help me get my head around this, or point me to another good resource. If I could see the method being implemented in a program or spreadsheet that would really help.

New Topic
This topic has been archived, and won't accept reply postings.
Loading Notifications...