Science, technology, health

Master Data Science

Objectives & future jobs :

This 2 year Master program aims at preparing high level scientist for research in a laboratory, possibly by pursuing a PhD program in data science, machine learning, signal processing, etc, or in a company as a research engineer.

All classes are given in English so that foreign students are welcome. A high level in mathematics (probability, statistics, optimization), computer science (algorithmics, programming, data bases) as well as machine learning (optional) is expected.

Most of the teachers are members of CRIStAL (computer science, automatics and signal processing) or of the laboratory Paul Painlevé (mathematics). Graduate students will be able to apply for the PhD programs of the best universities in France and abroad.

Programme

The structure will permit students to personalize their track by choosing appropriate courses, in particular in view of a subsequent PhD program.

Master 1

Each course is made of 24h in presence and typically represents 3 ECTS.

Master 2

Training opportunities :

This master is research oriented and prepares to a variety of professions related to an expertise in data science :

• Ph.D : in France or abroad.
• R&D : data scientists and machine learning engineers in companies or laboratories.
• Chief of project in AI and data science.
• Consulting : data science and innovation

Research support :

Associated research labs :

• CRIStAL (UMR 9189 CNRS)
• Laboratory of Mathematics Paul Painlevé (UMR 8524 CNRS)
• Inria Center of the University of Lille
• Research Center of IMT Nord Europe

Organisation of the training programme:

This master is made of 4 semesters.

M1: classes take place from the beginning of September to end of March. Then students experience an internship of at least 6 weeks and up to 18 weeks (4 ½ months), in a laboratory or a company.

M2: classes take place from the beginning of September to end of March. Then students experience an internship of at least 4 months (16 weeks) and up to 6 months, in a laboratory or a company.

Teachers are all experts in their field. They are all researchers in the field of data science and artificial intelligence or their applications.

Seminars are given by researchers who come to explain their area of interest at master level.

Reading groups are dedicated to a training to the data science & AI literature, as well as the culture of publication in research.

The data challenge takes place during a whole week that is dedicated to it, common to M1 and M2 and another AI & Data science of Centrale Lille, that is about 60 students altogether. The work is prepared in teams to solve some challenging problem, as well as to communicate on the proposed solution in a professional manner.

Research projects are prepared both in M1 and M2. This is an essential activity of the programme. About 130h of personal involvement is expected. The subject is proposed by a researcher from our associated labs. Students work individually from November to March on the project and produce a report as well as a presentation and the related code.

Examples of internships:

• Online graph inference for decentralized learning with heterogeneous data (Inria Lille, lab)
• Denoising astronomical images (LIRMM, Montpellier, lab)
• Automatic classification of work documents (Fondasol)
• Analysis of the quality of water (Véolia)
• Predicting decisions of justice (LexisNexis)
• Tracking the movement of table tennis balls (L3I/XLIM, Poitiers, lab)
• Prediction of flight delays (AirFrance)

Prerequisites and admission conditions

Access to Master 1

This master programme welcomes students with a strong background in mathematics and/or computer science at Bachelor level in a University or engineering school. A B2 level in English is required (a certificate of competence is expected).

This master is not open to “alternance”. With a company.

Access to Master 2

The registering in M2 is automatic after the success in M1 Data Science.

The external recruitment directly in M2 remains possible via the ecandidat.univ-lille.fr platform of the University of Lille. An excellent background in data science and machine learning on top of a high level profile in mathematics in necessary. A good knowledge/know-how of Python is required.

Since the number of places is limited, candidates are individually selected. The selection committee will put the emphasis on the past followed education program, the level of knowledge, internship reports, project reports, motivations, English level of each candidate.

Erasmus students are welcome.

Candidates are invited to send their application through the dedicated website. Information will be updated on the webpage: https://monmaster.gouv.fr