0
$\begingroup$

This is a general question. I understand the theory behind using multivariate Gaussian distributions, however one question still bugs me and I have not been able to get a "layman's" answer to it.

Given a new data point $X$, and a data set comprising various data points that can be classified into $N$ multivariate Gaussians. How is the point classified?

What I am asking is, given the $\{x_1,..,x_n\}$ dimensions that make up the data vector to be classified, how is the decision made whether that point is part of one Gaussian distribution vs. another?

$\endgroup$
1
$\begingroup$

Assuming you have N multivariate Gaussian PDFs $f_i(X)$, you could compare the PDF values like in maximum likelihood: $f_i(X)$ vs. $f_j(X)$, then pick the one with the highest value.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.