Comment on "A Declustering Criterion for Feature Extraction in Pattern Recognition

The purpose of this correspondence is to point out that the scaling procedure (as well as any nonsingular transformation) described in Section VI of the above paper1 is unnecessary.