.
This course presents the statistical methods exploited in data analysis
(factor analysis) or in modeling the explanatory relation of a response
variable (regression) and their use in the pyramid of business
intelligence.
The first part of the course is devoted to the factorial analysis which,
by confronting the representation spaces of individuals and variables,
enriches the interpretation and makes it possible to exhibit the
internal structure of the data. The data nature and coding lead to two
essential variants of factorial methods, namely Principal Component
Analysis (PCA) and Multiple Correspondence Analysis (MCA), combined in
Multiple Factor Analysis (MFA).
The second part presents different regression models and methods for
estimating their parameters, from the linear model to more complex
models with ill-known structure or suitable for unusual data
distribution.
- Responsable de cours: Atto, Abdourrahmane
- Responsable de cours: Couturier, Vincent
- Responsable de cours: Moliere, Emma