$$ \newcommand{\bx}{\boldsymbol{x}} \newcommand{\bt}{\boldsymbol{\theta}} \newcommand{\dkl}{\mathrm{d}_{\mathrm{KL}}} \newcommand{\dtv}{\mathrm{d}_{\mathrm{TV}}} \newcommand{\emv}{\hat{\theta}_{\mathrm{emv}}} \newcommand{\ent}{\mathrm{Ent}} $$

Lab sessions : Stochastic Control and Machine Learning

Author

Samy Mekkaoui

Published

February 3, 2026

Welcome to the webpage of the practical sessions of the course Machine learning and stochastic control taught by Professor Pham within the Master’s program in Probability and Finance (M2PF).

Global informations

Prerequisites :

  • Good knowledge on Probability Theory and Stochastic Processes.
  • Basics on Optimal control theory.
  • Basics on Deep Learning.
  • Familarity with Python and potentially with PyTorch.

Schedule :

  • Lab work n° \(1\) : Tuesday, January 27th, 9h-12h, room \(102\) in tower \(15-25\).
  • Lab work n° \(2\) : Tuesday, February 3rd, 9h-12h, room \(102\) in tower \(15-25\).
  • Lab work n° \(3\) : Tuesday, February 19th, 9h-12h, room \(106\) in tower \(14-15\).

Materials :

The practical sessions will be conducted in Python. You should bring your own laptop and have your own Python environment set up for each session.1

Planning :

  • Lab work n°1 : Deep PDE Solver

The first lab session will be about Deep PDE Solver for solving partial differential equations using neural networks. You will implement some of the algorithms seen during the course.

  • Lab work n°2 : RL for stochastic control problems

The second lab session will be on Reinforcement Learning for solving some stochastic control problems. We will implement some of the algorithms seen during the course.

  • Lab work n°3 : GenAI for data generation

The third lab session will be about Generative IA for data generation based on the Schrödinger Bridge. You will implement the Schrödinger Bridge Time series (SBTS) algorithm and apply it to generate some new samples of generic time series.

Grade :

You will have to choose one lab (TP) among the three proposed by modifying the column named “Lab Choice” associated to your name directly on the Excel sheet available here. Note that the project has to be done by groups of 2 or 3.

Your objective will be to answer the questions of the chosen lab, and an open-ended mini-project will be proposed at the end of it.

The final grade will be based on both the lab questions and the mini-project. In any case, the final course grade will be calculated as follows:

\[\begin{align} \text{Final Grade = 50 \% multiple-choice exam (QCM) + 50 \% (Lab +project)}. \end{align}\]

Questions ? :

If you have any questions, you can either e-mail Samy or Alexandre who prepared the lab sessions for this course.

Use this website

Structure of the website :

The site is structured into three sections that make up the course, each consisting of two chapters.2

  • Course reminders:

The first chapter, named Course reminders, reviews the key theoretical concepts covered in the course. It revisits important results and the corresponding algorithms for the underlying problems.

  • Practical session:

The second chapter, named Lab work, contains the practical session instructions along with a link to a Jupyter notebook where you can write your code. These notebooks are designed to be self-contained, allowing you to work independently during the sessions. At the end of each practical session, an open-ended project3 is provided. This project is not intended to be completed during the session, but it is closely related to the material covered and offers an opportunity to explore the concepts in more depth.

For your information, this site is generated using Quarto through GitHub Pages (see this GitHub page). If you spot any errors on the site, feel free to report them through pull requests.

Acknowledgements

We would like to thank the following individuals and organizations without whose support this course would not be possible:

  • Professor Huyên Pham for the support and valuable feedbacks during the creation of this lab.
  • Participants in the course for your interest in Machine Learning for Stochastic Control and its applications.

  1. The required scientific packages used during the practical sessions will be specified during the lab sessions.↩︎

  2. The sections will be uploaded shortly before the beginning of each lab session.↩︎

  3. The projects related to each course session will be uploaded after the third lab session, i.e., after February 19.↩︎