Can Pouliquen

Quick bio

Hello ! I am a 2nd year PhD student in Machine Learning at the Ecole Normale Supérieure de Lyon under the supervision of Mathurin Massias, Titouan Vayer and Paulo Gonçalves. I am a part of OCKHAM, a joint team of Inria¹ and CNRS², led by Rémi Gribonval.

Previously, I was a research intern in Deep Learning in the NeuroAI team of CERCO, a CNRS lab in Toulouse³, under the supervision of Andrea Alamia and Rufin VanRullen, where I had the opportunity to work on the oscillatory dynamics of artificial neural networks under some biological feasibility constraints.

Before that, we may have met in Berlin, Montpellier, Lyon, Istanbul, Beyrouth, Amman or Cairo. Hit me up if that's the case !

Outside of research, I enjoy hiking, running, food, beers, and "la belle vie" with my girlfriend⁴.

Research

I am currently working on large-scale optimization for statistical machine learning in the context of sparse graph learning.

A bit more formally, given an empirical covariance matrix \(S \in \mathbb{R}^{p \times p}\) of a dataset of \(n\) random vectors \( x_i \sim \mathcal{N}^p(\mu, \Sigma) \), I am interested in estimating \( \Theta = \Sigma^{-1}\) (called precision matrix) with a sparsity prior. I am currently trying to bridge the gap between convex optimization-based approaches and deep learning-based approaches to do so.

I have a particular interest in applications to neuroscience problems.

Here is a list of links to my publications:

• Schur's Positive-Definite Network: Deep Learning in the SPD cone with structure.
  Can Pouliquen, Mathurin Massias, Titouan Vayer
  Preprint, 2024.


• Implicit differentiation for hyperparameter tuning the weighted Graphical Lasso.
  Can Pouliquen, Paulo Gonçalves, Mathurin Massias, Titouan Vayer
  GRETSI, 2023.


• Can sparsity improve the privacy of neural networks ?
  Antoine Gonon, Léon Zheng, Clément Lalanne, Quoc-Tung Le, Guillaume Lauga, Can Pouliquen
  GRETSI, 2023.

Teaching

ENS de Lyon: Convex optimization M1 (2023-2024)
TA of Mathurin Massias, 12h

CPE Lyon: Optimization algorithms M1 (2023-2024)
TA of Marion Foare, 20h

Ecole Centrale de Lyon: Probability theory L3 (2023-2024)
TA of Céline Helbert and Christophette Blanchet, 32h

Talks

• Deep Learning in the SPD cone with structure.
  Poster session at the CIRM, June 2024.
  Poster session at SMAI-MODE, March 2024.


• Machine Learning, Optimization, and interactions.
  Scientific outreach for an audience of high-school students, June 2024.


• Implicit differentiation for hyperparameter tuning the weighted Graphical Lasso.
  GRETSI oral, August 2023.


• Neuroscience inspired machine learning.
  OCKHAM seminar, June 2023.


• Estimating dynamic functional connectivity graphs.
  LIP PhD seminar, May 2023.

Contact

To get in touch:

• Email: can.pouliquen \AT\ ens-lyon \DOT\ fr
• Office: M7-1H29, ENS Lyon, site Monod, 46 allée d'Italie, 69007, Lyon, France
• Tel: If you know me already, shoot me an email so I can send it to you :)