Abstract
This project has for aim to classify the population into
socio-economic categories. In order to cluster the population into homogenous
classes, we have been based on different variables or criteria. We distinguish
between two kind of criteria :
· socio-economic criteria like principal occupation,
socioprofessional status, high diploma...etc.
· Socio-demographic criteria such us age, sex,
matrimonial status, family relationship with householder... etc.
So as to manage this typology of population, we use two famous
statistic methods : hierarchical clustering and k-means method. We applied
those methods using data stem from household's standard of living
survey(1990/1991). We distinguish between two environment : rural and urban.
This distinction was justified by the social disparities between city dwellers
and country people.
The results are as good as the same between the two
populations expect some categories which exist in urban environment and not in
rural one such us the category of senior or middle executive in administration
and senior executive having scientific and liberal profession.
Key words : hierarchical cluster, k-means
method, multiple correspondence analysis
Table des matières
Page
Liste des
abréviations........................................................................................VII
Liste des tableaux et
figures...............................................................................IX
1 Introduction
générale
....................................................................... 10
| 
