search for


New link of multiple correspondence analysis and K-means cluster analysis
Journal of the Korean Data & Information Science Society 2022;33:1043-52
Published online November 30, 2022;
© 2022 Korean Data and Information Science Society.

Dong Hyun Kim1 · Gyemin Lee2

12Department of information and Statistics, Gyessang National University
Correspondence to: 1 Graduate student, Department of Information and Statistics, Gyeongsang National University, Jinju 52828, Korea.
2 Professor, Department of information and Statistics, Gyeongsang National University, Jinju 52828, Korea. E-mail:
Received October 25, 2022; Revised November 9, 2022; Accepted November 11, 2022.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Multiple correspondence analysis (MCA) for categorical data is a good tool to explore the relationship between categorical variables, but it is not easy to investigate whether heterogeneous groups exist between subjects. In this paper, a new method combining MCA and K-means cluster analysis (CA) was proposed to classify subjects into several similar groups. Through a simulation, we compared the proposed method with several existing methods that combine MCA and CA, a method that sequentially applied MCA and CA, and a method that uses only CA. As a result of the simulation study, the proposed method showed superior or similar performance compared to other methods with respect to the performance of finding real clusters.
Keywords : lternating least square, dimension reduction, multiple correspondence analysis, K-means cluster analysis.