Whenever we talk about the probability distribution of data having more than one feature, we have only one option i-e multivariate normal distribution. Is there any other probability distribution model exist for multivariate data? If yes, then how can we find its parameters using MLE in MATLAB
$\begingroup$
$\endgroup$
1
-
$\begingroup$ It's worth noting that the phrasing of this is incorrect (not to be an asshole). For linear regression models, we assume the data comes from a multivariate normal distribution, then we estimate the parameters of that distribution (i.e., the covariance matrix and mean vector). $\endgroup$– franciscojavierarceoNov 18, 2016 at 19:42
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
There are quite a number of multivariate distributions, in theory. At the outset, we categorize them as multivariate discrete distributions and multivariate continuous distributions.
To start with, you may explore the following link for : multivariate discrete distributions and the following for multivariate continuous distributions.
There are entire volumes printed on the subject, for example, books by: N. Balakrishnan, Norman L Johnson, and Samuel Kotz and these books are published by John Wiley and Sons, Inc.
