[go: up one dir, main page]
photo de lenaic 

Lénaïc Chizat

I am a tenure track assistant professor at EPFL in the Institute of Mathematics, where I lead the DOLA chair (Dynamics Of Learning Algorithms).

Main research topics:

  • Continuous Optimization: gradient flows, algorithmic regularization, space of measures, interacting particle methods

  • Optimal Transport: extensions, statistical aspects, entropic regularization, applications to machine learning

  • Theory of deep learning and neural networks

To contact me : firstname.lastname@epfl.ch

Publications

[Google Scholar profile]

Curriculum Vitae

[PDF]

Master courses

Some resources

  • Invited blog post on Francis Bach's research blog, an overview of our recent results on the predictor learnt by two-layer neural networks.

  • Slides for the best paper award lecture at SIAM-IP 2022 about my research on sparse optimization on measures with overparamterized gradient descent.

  • Slides about estimating optimal transport with entropic regularization

  • Slides about infinitely wide two-layer neural networks

  • Slides about Trajectory Inference with Mean-Field Langevin Dynamics

  • All the numerical experiments in my papers are reproducible using the codes I post on Github

PhD students and Post-docs

Miscellaneous