Mixed-Condorcet: An Advanced Method of Clustering for Mixed Data in Social Sciences

Abstract

This paper provides a new clustering method for mixed data based on α-Condorcet, denoted mixed-Condorcet, by introducing a new Condorcet criterion. This criterion combines α-Condorcet and k-prototype criteria. Next, we give the within-cluster sum-of-squares expression for our new method. Furthermore, we compare mixed-Condorcet clustering with k-prototype and Kamila clustering. The comparison employs quality index (QI) and a within cluster sum of squares index. Our findings are illustrated using both simulated and real datasets.