Céline Moucer

INRIA SIERRA Team, Ecole normale supérieure, Paris France.

prof_pic_2.jpg

2 rue Simone Iff

75012 Paris, France

Office C406 (on floor 4)

celine[dot]moucer[at]gmail[dot]com

Hello, and welcome to my website!

I am Céline Moucer, currently working at the French Ministry of Finance (Direction du Budget). Since taking on this role, I have slowed down my research activities, but I’m always happy to receive suggestions, comments, and questions, and I make sure to respond in a timely manner.

In october 2024, I completed a PhD in optimization for machine learning, under the supervision of Adrien Taylor and Francis Bach. My research focuses on developing constructive approaches to concentration inequalities and first-order convex optimization methods, exploring their applications to potentially sparse machine learning problems and their continuous-time formulations. I am particularly interested in computer-aided techniques for the worst-case analysis of optimization methods, also known as Performance Estimation Problems.

Before starting my PhD, I graduated from Ecole Polytechnique (X2016) and got a master degree in Maths, Vision and Learning in 2020 (master MVA) from Ecole Normale Supérieure and Ecole des Ponts. I am also a civil servant engineer (Corps des IPEF), working for the French administration.

You can find a list of my publications and codes here.

news

Oct 1, 2024 I have successfully defended my PhD thesis at Inria Paris!
Aug 30, 2023 Our paper on continuous-time Lyapunov analysis was accepted at SIOPT!
Aug 24, 2023 Presenting PEPit and continuous-time PEP at EUROPT2023.
Feb 14, 2023 PEP talks took place à Louvain-la-Neuve, thanks to all the participants!
Oct 24, 2022 During two days, we welcome 25 high school girls at INRIA Paris as part of the RJMI program.
Jul 25, 2022 Presenting PEPit at ICCOPT.
Sep 1, 2021 Starting my PhD at INRIA Paris.

selected publications

  1. journal
    PEPit: computer-assisted worst-case analyses of first-order optimization methods in Python
    Baptiste Goujaud, Céline Moucer, François Glineur, Julien Hendrickx, Adrien Taylor, and Aymeric Dieuleveut
    Mathematical Programming Computation 2024
  2. journal
    A systematic approach to Lyapunov analyses of continuous-time models in convex optimization
    Céline Moucer, Adrien Taylor, and Francis Bach
    SIAM journal on Optimization 2023
  3. preprint
    Geometry-dependent matching pursuit: a transition phase for convergence on linear regression and LASSO
    Céline Moucer, Adrien Taylor, and Francis Bach
    2024
  4. preprint
    Constructive approaches to concentration inequalities with independent random variables
    Celine MoucerAdrien Taylor, and Francis Bach
    2024