Goals

The objective of the course is to introduce the main elements of the theory of linear and non-linear regression. We will study the theoretical foundations of regression, as well as the fundamental questions to ask when modeling real-world phenomena.

Content

The main concepts studied are as follows: Normal, Chi-square, Student and Fisher-Snedecor distributions (as well as the relationships between them), random vectors and more particularly Gaussian random vectors, simple linear regression, multiple linear regression, multicollinearity, diagnostics, choice of linear model and non-linear regression.

Teaching methods

The course material will be presented during ex-cathedra lectures. The slides used will be available on Webcampus. 

The exercise sessions will mainly focus on learning the R statistical software and applying regression methods to real data sets.

Assessment method

The course assessment will consist of two parts:

  1. A written exam during the exam session, with questions focusing on knowledge and understanding of the material covered in class and during the exercise sessions.
  2. Continuous assessment. This will consist of various assignments to be completed by students; these assignments may be subject to one or more written and/or oral assessments, depending on the number of students enrolled. The assignments may include the analysis of real data sets and may cover the following topics: modeling, model parameter estimation, variable selection, model validity testing, etc.

The written exam during the exam session and the continuous assessment will be organized according to the instructions provided by the instructor during the course sessions. 

Language of instruction

French
Training Study programme Block Credits Mandatory
Bachelor in Mathematics Standard 0 5
Bachelor in Mathematics Standard 3 5