A Note On Some Statistical Aspects of “Speaker-Independent Recognition of Isolated Words Using Clustering Techniques”