PSL Preparatory Weeks

Preparation courses on machine learning, 2021

This year, the transverse program for data sciences of PSL proposes two preparatory weeks :

  • The preparatory week on basics in math and data science (starting the 30th of August). The students will follow it remotely and autonomously. For each part, video lectures, exercise sheets and lab sessions (Python notebooks) will be provided.
  • The preparatory week on machine learning, bigdata, and ethics (6 to 11 Sept. 2021)

Both weeks are open to all master and PhD students from PSL. Subject to availabilities, students from other universities can also attend. The registration is free but mandatory. See the two links below, one for each week.

Preparatory week on basics

This first preparatory week will start on Monday, August 30. The students will follow it remotely and autonomously. For each of the five parts, video lectures, exercise sheets and lab sessions under the form of Python notebooks will be provided. For each part, the theoretical content (lectures and exercises) should represent about a half day of work (around four hours), and the lab session also. It is advised to watch the theoretical content in the morning, and do the lab session in the afternoon of the same day.


You can pre-register here to the course.

PSL students have priority if they pre-register before August 15th.

Expected roadmap

With this organization in mind, the schedule is as follows :

  • August 30 (Reza Hatami) Functions and sequences
  • August 31 (Gwendoline de Bie) Basics of linear algebra
  • September 1 (Pierre Ablin) Differential calculus and PCA
  • September 2 (Madalina Olteanu) Statistics
  • September 3 (Pierre Senellart) Databases

Q&A sessions

In addition, sessions of questions and answers will be organized : the teachers of the first four parts will make themselves available through Teams in the afternoon, during two hours, and will answer the questions of the students.

Preparatory week on machine learning, bigdata and ethics

Dates and (expected) location

Dates: 6 to 11 Sept. 2021.

Location: Dauphine + online (TBC).

The sessions will be both at Dauphine (if possible) and remotely.


You can pre-register here to the course.

PSL students have priority if they pre-register before August 15th.



  • 9:00–10:30: (course) Machine learning: recent successes.
  • 11:00-12:30: (course) Introduction to machine learning.
  • 14:00-17:00: (lab session) Introduction to Python and Numpy for data sciences.


  • 9:00–10:30: (course) Machine learning models (linear, trees, neural networks).
  • 11:00-12:30: (course) Scikit-learn: estimation/prediction/transformation.
  • 14:00-17:00: (lab session) Practice of Scikit-learn.


  • 9:00–10:30: (course) Optimization for linear models.
  • 11:00-12:30: (course) Practical optimization for ML.
  • 14:00-17:00: (lab session) Logistic regression with gradient descent.


  • 9:00–10:30: (course) Introduction to deep learning.
  • 11:00-12:30: (course) Differentiable programming.
  • 14:00-17:00: (lab session) Classification with PyTorch and GPUs.


  • 10:00-13:00: (course) Introduction to spark for ML 1/2
  • 14:00-17:00: (lab session) spark for ML


  • 9:00-12:00: (course) Ethic in Data Science

Course materials

Slides and suplementary materials for the lecture and numerics are available from here (comming soon)

Practical information

The afternoon are dedicated to practical sessions using Python. Students will be on their own under a weak supervision from the teachers. Students can use the discord server chat to communicate, share information, codes, data and help each other during these session.

These practical sessions will necessitate the use of Python 3 with the standard Scipy ecosystem, Scikit-learn and Pytorch. They will make use of Jupyter notebooks. The easiest way to proceed is to have a gmail account and make use of a remote Google Colab to run the notebooks. If you are not confortable with this, or if you want to run the code locally, you need to install the requested python package, preferably using Anaconda.