ENSAE Paris - École d'ingénieurs pour l'économie, la data science, la finance et l'actuariat

Data Science, Statistics and Learning

Data Science, Statistics and Learning


The third year of the ENSAE Paris engineering program includes six specialization tracks. Each track has been designed to offer a coherent sequence of courses (advanced theoretical courses, applications, projects, seminars, etc.), preparing students for a major field of work and giving them a broad and in-depth vision of that field. Each of the professions practiced by ENSAEs will call to varying degrees on statistical methods, data science and modeling, which are therefore present within each track.

Learn more about admissions to the engineering program

The "Data science, statistics & learning" track offers fundamental training in statistics, machine learning and more generally in data science and artificial intelligence.

Professors in charge of the track

Alexandre Tsybakov

Vianney Perchet


Using the most appropriate computer tools, this track trains "data scientists" with a very high level of scientific expertise, with openings to the most active application fields (finance, insurance, social sciences and possibly basic skills in biology).

This pedagogical approach, based on a solid theoretical mathematical foundation, ensures a better assimilation of knowledge, a wise use of algorithms, and encourages creativity and innovation.


The "Data science, statistics and learning" track aims to deliver both broad and deep skills in designing statistical models, developing artificial intelligence algorithms, and organizing testing or statistical learning to support decision making in bounded rationality.

The courses oriented towards learning and large-scale statistics lead to jobs as statistical experts in industry (EDF, Air Liquide, Thales, etc.), in companies using large databases (Google, Apple, Facebook, Amazon and Microsoft), but also in the finance and insurance sectors (AXA, BNP Paris, CFM, etc.) or in technology start-ups

The courses related to statistical surveys lead to jobs as methodologists in survey institutes, in the statistics and research departments of large companies and government agencies, and in consulting firms. Supported by fundamental theoretical courses, this path also leads to research in statistics and machine learning.


The compulsory scientific courses and the optional courses recommended for the track are described below. Each semester, you can choose one modern language (maximum). English is compulsory if your level is lower than B2. The options can be mixed between the different tracks (under time constraints) to create hybrid profiles. It is then recommended to discuss their coherence, as well as the articulation of the choice of courses with a possible M2 followed in parallel, with the Master studies director or/and the person in charge of the tracks.

First semester

Course ECTS Hours (course+tutorials)
Advanced Machine Learning 4 21+9
Bayesian Statistics 3 18+0
Ethics and responsibility in data science 2 12+0

You can choose from 3 to 7 options (including foreign language) from the entire 3A course catalog, semester 1 (subject to compatibility with your schedule), to reach a total of 30 to 31 ECTS for the semester. We recommend the following courses in this track:

Course ECTS Hours (course+tutorials)
Advanced Econometrics: Panel data and duration models 4 24+0
Algorithm Design and Analysis 3 18+0
Blockchain: Bitcoin and Smart-Contracts 3 18+0
Entrepreneurship 1 3 18+0
Estimation non paramétrique 4 15+9
Hidden Markov models and Sequential Monte-Carlo Methods 3 18+0
Hi!ckathon 2 0+0
High-dimensional statistics 4 15+9
Modeling and managing energy risks 2 12+0
Advanced Optimisation 4 24+0
Software Infrastructures and Systems 3 18+0
Statistical methods of Econometrics 3 18+0

Second semester

Course ECTS Hours (coursr+tutorials)
Machine Learning for Natural Language Processing 3 18+0
Analysis of Matrix Data 4 21+0

You can choose from 4 to 6 options (including foreign language) from the entire 3A course catalog, semester 1 (subject to compatibility with your schedule), to reach a total of 30 to 31 ECTS over the semester. We recommend the following courses in this track:

Course ECTS Hours (course+tutorials)
Bootstrap and Resampling Methods 3 18+0
Data Storytelling 3 18+0
Deep Learning: Models and Optimization 3 18+0
Deployment of Data-Science Projects 2 12+0
Digital Entrepreneurship 3 18+0
Entrepreneurship 2 3 18+0
Fairness and privacy in machine learning 3 18+0
GPU program 2 18+0
History and epistemology of Statistics 3 18+0
Machine learning for Econometrics 4 24+0
Online Learning and Aggregation 3 15+6
Optimal Transport : from Theory to Tweaks, Computations and Applications in Machine Learning 3 12+6
Parallel Programming for Machine Learning 3 18+0
Reinforcement learning 3 18+0
Sampling Methods: From MCMC to Generative Modelling 3 12+6    
Statistics 3 4 24+0
Statistical Optimal Transport 2 12+6