Clément Lezane
Postdoctoral Researcher in Optimization
📍 [ANITI Toulouse]
📧 [clement.lezane@univ-toulouse.fr]
| Publications | Talks/Slides | Code/Engineering Project |
About Me
I am a Postdoctoral Researcher at ANITI Toulouse, working in collaboration with Jérôme Bolte, Jean-Michel Loubes, and François Bachoc. My current research focuses on developing rigorous machine learning metrics to evaluate and benchmark the fairness of existing optimization algorithms.
Prior to joining ANITI, I earned my Ph.D. in Applied Mathematics from the University of Twente (Department of Stochastic Operations Research). My doctoral research, co-supervised by Sophie Langer, Wouter M. Koolen, Cristóbal Guzmán, and Alexandre d’Aspremont, culminated in my thesis, “First Order Methods with Non-Euclidean Geometry.” During my doctoral studies, I collaborated closely with the SIERRA team at Inria Paris to develop accelerated optimization methods for machine learning applications.
Research Interests
My research lies at the intersection of optimization, machine learning, and statistics, with a focus on designing optimal, accelerated algorithms for complex, large-scale problems. My specific areas of focus include:
- Stochastic Convex Optimization: Complementary composite minimization and randomized smoothing.
- Accelerated First-Order Methods: Mirror descent, non-Euclidean geometry, and star-convex functions.
- Imbalanced Learning: Algorithm design and analysis for datasets with highly skewed class distributions.
- Minimax Problems: Theoretical foundations and applications in algorithmic fairness and differential privacy.