If you have any questions, comments, etc. please post them on this page.

Go to Discussion about Discriminant Functions for the Multivariate Normal Density .

Questions and Comments

Reviewed by Raj Praveen

This slecture discusses about the various forms of discriminant functions for multivariate normal data based on the covariance matrix.

The order in which the slecture discusses the discriminant functions makes the topic very easy to follow. The author breaks down the mathematical components of the multivariate density function and explains the importance of the covariance matrix. The slecture then proceeds with examining the discriminant function in three cases with respect to the covariance matrix. In each case, the form of the classifier obtained (linear or quadratic) and the corresponding decision boundary has been mathematically explained well. The slecture also includes relevant figures to illustrate the form of the classifiers. It would have been even better, if the example 2-D data discussed had been plugged in the discriminant functions and solved practically to illustrate the classifiers obtained.

On a whole it is a very well explained and illustrated slecture.

Comments by Yanzhe Cui

Raj, thanks for your review comments. I tried my best to make this slecture to easily follow. I will make my slecture better.