Problem $4G$ only $-$ Correlation$/$covariance derived information. $\{16\}$ Suppose we know

that the covariance matrix $$ of attributes is identical to the correlation matrix $($i$.$e$.$ matrix whose

$(i,j)$ entry gives the correlation between attributes $x_i$ and $\{$:$x_j).$ What can we say about the relation

between these attributes? Are there more important and less important attributes? Assume that

no attribute have the same values in all records, variance values of attributes differ widely and that

all information is contained in attribute variances.