Given a set X of n feature vectors x 1= (x 11,…,x 1p ), ..., x n = (x n1,…,x np ) of dimension p, the problem is to compute the sample means and variance-covariance matrix or correlation matrix:
|
Statistic |
Definition |
|---|---|
|
Means |
M = (m(1), ..., m(p)), where
|
|
Variance-covariance matrix |
Cov = (v i j ), where
|
|
Correlation matrix |
Cor =
(c
i j
), where
|
