Dispersion matrix
A measure of Linearity relationship
Variance is for 1-dimensional data while Covariance Matrix is for vector data distribution
Definition
using definition
Pairwise Independence → Covariance 0, not reversal. But jointly normally distributed but uncorrelated, then they are indeed independent.
Since covariance is the expected value of the product of deviations from their respective means for X and Y, it can be interpreted as the dot product of two vectors. In particular, Correlation is like cosine, bounded between -1 and 1, as it's equivalent to dividing by the vector magnitudes.