Skip to main content

Descriptive sheet of the TU MET-MALE

TU MET-MALE - Methods of machine learning

Objective

Understanding and mastering the main methods of machine learning to analyse health data.

Knowledge to be acquired

Know the main methods of machine learning, from their fundamental concepts to their contextualised use to analyse health data.

Skills to be acquired

Being able to perform and interpret an analysis using machine learning method.

Description

  • General theoretical framework for machine learning.
  • k-means.
  • Support vector machine.
  • Tree-based methods, baggin, random forests, boosting.
  • Neural networks.

Pedagogical method

Courses on-line by NetConference, distance mentoring over the Internet.

Pre-requisite

Knowledge of computer use, word processing, computer presentation tools.

Total hours

Between 20 and 30 hours per NetConference + personal work.

ECTS

6

Place

Faculté des sciences médicales et paramédicale de Marseille

Dates

See the schedule for this TU 

Head

R. Giorgi

Teachers

R. Giorgi, Q. Marcou, N. Ngo.

TU – mandatory

This TU is mandatory for the Master 2 of Public Health:

  • AI4PH: "Artificial Intelligence for Public Health"

TU – optional

 

Assessment of knowledge

A single examination is organised, in accordance with the Aix Marseille University rule.

For this single session, the assessment of knowledge of this teaching unit is based on:

  • Continuous assessment of knowledge with various homework assignments, tests and exercises. This continuous assessment will have a coefficient of 2 in the final mark.
  • An oral examination. This exam will have a coefficient of 3 in the final grade. A personal convocation will be sent by e-mail to each student.

To validate a UE, it is necessary to obtain a mark ≥ 10.

To validate the DESU, a single final mark ≥ 10 must be obtained, knowing that there is no compensation between the UEs.

Educational resources

Course materials on the Internet.