Comments on "On the selection of the dominant poles of a system to be retained in a low-order model"

A limitation of the model order selection criterion as proposed in the above correspondence is pointed out. It is shown that the results obtained from the criterion are not generally unique and acceptable.