Colinéarité et régression linéaire
The linear analysis of the regression, called also more simply linear regression, is one of the most used statistical methods in applied sciences and social sciences. Its objective is double: first of all it consists in describing the relations between a variable, called explained (or dependent) var...
Gorde:
| Argitaratua izan da: | URI:https://journals.openedition.org/msh, |
|---|---|
| Egile nagusia: | |
| Formatua: | Article ou chapitre numérique |
| Hizkuntza: | Français |
| Argitaratua: |
Mathématiques et sciences humaines
2006
|
| Gaiak: | |
| Sarrera elektronikoa: | Accès Université d'Orléans et IFPM Accès Université d'Orléans et IFPM |
| Gaia: | The linear analysis of the regression, called also more simply linear regression, is one of the most used statistical methods in applied sciences and social sciences. Its objective is double: first of all it consists in describing the relations between a variable, called explained (or dependent) variable, and several variables, called explanatory (or independent) variables. It also makes it possible to conduct forecasts of the explained variable in terms of the explanatory variables. The links between the explanatory variables exert a considerable influence on the effectiveness of the method, whatever the objective in which it is used. We expose in this paper some of the properties of these links, recently proved and published in several papers |
|---|