Computer Programs for Hierarchical Polythetic Classification ("Similarity Analyses")

Open Access
Publisher Website
Google Scholar
Abstract
It is demonstrated that hierarchical classifications have computational advantages over cluster analyses. Flexible programs are outlined providing two sorting strategies and four alternative similarity-coefficients. Preliminary results suggest that for qualitative data one of the strategies and two of the coefficients are superior to the remainder of the system; the further refinements desirable for a large-scale production program are discussed.

This publication has 0 references indexed in Scilit: