return the Mahalanobis' D-square distance
d(k) = (X(k,:)-MU)*inv(SIGMA)*(X(k,:)-MU)'
where MU and SIGMA are the mean and the covariance matrix of Y
return the Mahalanobis' D-square distance between the multivariate samples x and y, which must have the same number of components (columns), but may have a different number of observations (rows).