An enduring interest in classification: Supervised and unsupervised

Abstract

I have researched in the field of discriminant analysis for over 40 years and for nearly as long in the field of cluster analysis. Thus, I think it is fair to say that I have had an enduring interest in discriminant and cluster analyses, that is, in classification both supervised and unsupervised. The latter terminology is used outside of statistics in fields such as artificial intelligence, machine learning, and pattern recognition. However, the gap between these fields and statistics has narrowed appreciably over the years, and discriminant analysis and cluster analysis are also often referred in statistics as supervised classification and unsupervised classification, respectively.