Multivariate Distance Between Data Structures Determined By Multiple Correspondence Analysis.

Authors

  • Mawency Vergel Universidad Francisco de Paula Santander, Cúcuta, Colombia
  • Maura Vásquez Universidad Central de Venezuela
  • Guillermo Ramírez Universidad Central de Venezuela

DOI:

https://doi.org/10.61799/2216-0388.2336

Keywords:

Multiple correspondence analysis, STATIS, Hilbert Schmidt distance, comparison between occasions.

Abstract

This article focuses on a statistical procedure called DIST-ACM, which aims to perform a comparative analysis of data structures determined by measurements taken on the same individuals, characterized according to the same set of categorical variables on different occasions. Multiple correspondence analysis (MCA) is applied on each occasion. The comparative analysis of the data structures generated on these different occasions is central to the problem addressed. This is accomplished through the design, testing, and application of a measure of the multivariate differences between the data structures obtained on each occasion. The DIST-ACM measure, based on the Hilbert-Schmidt distance, follows the approach defined in the STATIS methodology, in which the same individuals are described by a group of continuous variables evaluated on K occasions, but adapted here to the treatment of categorical data. The research has focused on measuring changes in individual behavior, influenced by the same set of categorical variables, over time. The proposed procedure has been applied to a set of real data related to the teaching and learning process of subjects that focus on the modeling and analysis of dynamic systems, differential equations, and numerical methods, the first taking precedence over the second.

Downloads

Download data is not yet available.

References

X

Published

2026-09-01

Issue

Section

Artículo Originales

How to Cite

[1]
Vergel, M. et al. 2026. Multivariate Distance Between Data Structures Determined By Multiple Correspondence Analysis. Mundo FESC Journal. 16, 36 (Sep. 2026). DOI:https://doi.org/10.61799/2216-0388.2336.