Exploratory data analysis: an application on a medical topic

Authors

  • Marie-Paule Kestemont Université Catholique de Louvain

DOI:

https://doi.org/10.6092/issn.1973-2201/669

Abstract

A model of exploratory data analysis techniques is presented in this paper. The starting point of the study is a data set for which the interactions between its elements are unknown. The objective is to exhibit these interactions -if they exist-and to define some structures of relation between different factors We want to isolate from the data set the main factors by the method of principal components. Once these leading directions are obtained, we try to establish the relation of the individuals to these directions, without however defining a link of causality (the individuals are grouped together according to some habits and some characteristics). The principal components analysis is a descriptive analysis. The different groups only present individuals with respect to other within the framework of the principal factors we have found. The data are of medical kind. The analysis of those data shows three head directions :-a general factor of cardiac and vascular risk, -a factor of "acids", - a psychic factor. In the framework of those factors, we compare the positions of individuals according to their habits and some other characteristics (smoke habit, food habit, blood-group).

Published

2008-04-16

How to Cite

Kestemont, M.-P. (1985). Exploratory data analysis: an application on a medical topic. Statistica, 45(1), 33–35. https://doi.org/10.6092/issn.1973-2201/669

Issue

Section

Articles