Statistical learning

This course deals with the machine learning framework from a statistical point of view.
We will focus mainly on the supervised framework (regression and classification) and introduce some elements of the unsupervised framework through partitioning methods (clustering).
In addition to modeling and theoretical aspects, the course will also cover some elements of optimization and implementation (sklearn, pytorch, etc.) of the methods introduced.

Be able to model a new learning problem in the light of the objectives and methods available.

Linear model (HAX814X) / Software development (HAX712X) / Optimization (HAX706X)

 
 
Recommended prerequisites: Inferential statistics (HAX710X) and Estimation and nonparametric tests (HAX809X)

CC

Graded practical work (code)

Project (report + presentation + code)

• Introduction to supervised learning; linear models.
• Cross-validation, logistic regression, discriminant analysis.
• Model selection and regularization methods.
• Performance measurement (multi-class: top-k, ROC curve AUC, etc.)
• Perceptron and stochastic gradient (descent).
• SVM
• Decision trees, Random forests and Boosting.
• Unsupervised learning (partitioning: KMEANS, Ward's method)
• Neural networks 

 
Timetable:
CM: 21h
TD:
TP:
Field :