Statistique exploratoire multi modèle

Objectifs

-       With the explosion of big data problems statistical learning has become a very hot field. In this course, many linear and non-linear statistical methods are discussed and practiced. Teaching is resolutely focused on practice with R or Python practical works for each method (20% theory, 80% practice).

-       Students will be able to optimize each model to compare them and ultimately select the most efficient method on the available data.

Description

-       Lesson + practical work for each part :

Introduction : statistical learning, regression & classification – Linear models - GAM – Decision trees – Model aggregation methods (Bagging, Random forests, Boosting) – Support Vector Machines – Neural Networks & Deep Learning

Bibliographie

-       An introduction to statistical learning, G.James & al., Springer

-       The elements of statistical learning, T.Hastie & al., Springer

-       https://cran.r-project.org/

Conditions d'admission

-       R & Python, inferential statistics, gaussian linear model, logistic regression

Session 1 ou session unique - Contrôle des connaissances

Session 2 - Contrôle des connaissances