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

Key Points

• The aim is to develop and validate a novel clustering method called mixed-Condorcet for mixed data analysis.
• Introduced a new Condorcet criterion combining α-Condorcet and k-prototype methods.
• Derived the within-cluster sum-of-squares expression for mixed-Condorcet.
• Compared mixed-Condorcet with k-prototype and Kamila clustering using quality index and within-cluster sum of squares.
• Mixed-Condorcet shows improved clustering accuracy compared to k-prototype and Kamila methods.
• Simulation and real datasets provided evidence supporting the efficacy of the mixed-Condorcet approach.

Abstract