Paul Mangold

Presentation

Since November 2023, I am a post-doctoral researcher at Ecole Polytechnique, supervised by Eric Moulines.

Before Paris, I did a PhD thesis titled Exploiting Problem Structure in Privacy-Preserving Optimization and Machine Learning, in the Magnet team at Inria Lille. My supervisors were are Aurélien Bellet, Marc Tommasi, and Joseph Salmon. During my PhD, I also worked with Michaël Perrot and Hadrien Hendrikx.

Before Lille, I was student at ENS de Lyon. I followed the Master Datasciences at Université Paris-Saclay.

In 2019, I did the agrégation de mathématiques, option informatique. This page (in french) contains all documents related to this: lessons, proofs, together with a few resources, links and remarks.

Research Interests

My research is centered around optimization, federated (reinforcement) learning, and ethical concerns that come with training machine learning models. Quite generally, I am interested in finding algorithms that can adapt to problem's structural properties to improve either convergence rate, communication complexity, or other properties like privacy and fairness.

I am also strongly interested in methods that make training and using models more practical and trustworthy. In particular in quantification of model's uncertainty, and in procedures that allow learning with missing data (e.g. imputation).

As of now, I have mostly studied convex problems, but I hope to work on more general classes of non-convex problems soon!

Feel free (and even, encouraged) to reach to me if you want to discuss any of these questions, I am always happy to chat!

You can find a list of my publications on the dedicated page or on Google scholar, and you can contact me at paul.mangold polytechnique edu. I am also on Twitter.

News

• February 2024. New preprint online. We study federated linear stochastic approximation, propose a new method that uses control variates to mitigate bias, and apply it to federated temporal difference learning.
• January 2024. Our paper has been accepted to AISTATS!
• November 2023. I am starting a post-doc at Ecole Polytechnique.
• October 2023. I just defended my thesis! :)
• August 2023. New preprint online! We propose a variant of the Gaussian Mechanism that can exploit a relaxed sensitivity assumption, and apply it to private gradient descent.
• July 2023. I finished writing my thesis! It was intense but I am happy with the result.
• July 2023. I attended the CAp conference at Strasbourg. CAp is a nice conference for the francophone community to meet.
• April 2023. Paper accepted to ICML 2023!
• April 2023. I am in Valencia for AISTATS 2023 to present our paper [and our poster] on high dimensional private empirical risk minimization using a greedy coordinate descent approach.
• April 2023. I am presenting a poster on the interplay of fairness and privacy at the statlearn workshop in Montpellier.

Teaching

I teach at Lille University, where I do a machine learning course for master's students and a machine learning/graph cours for bachelor students. More details on this page.

Tools I use

I am a fervent user of emacs, and am starting to use org-mode and org-ref, that are really fantastic tools.

I generally code in C++ and python using numpy, sci-kit learn and numba. I also use tuna for profiling, which is a really nice visualisation tool.

You can find the templates of my posters and presentations on the following repository.