Florian Schäfer

S1317 CODA

756 W Peachtree St

Atlanta, GA 30332

I’m Florian Schäfer, assistant professor in the School of Computational Science and Engineering at Georgia Tech.

Prior to joining Georgia Tech, I received my PhD in applied and computational mathematics at Caltech, working with Houman Owhadi. Before that, I received my Bachelor’s and Master’s degrees in Mathematics at the University of Bonn.

My research interests lie at the interface of numerical computation, statistical inference, and competitive games. To learn more, have a look at these high-level summaries or the defense of my thesis.

I am currently hiring PhD students. If you have a strong background in computational mathematics and the above seems interesting (or you have a project in mind that I might find interesting), feel free to reach out!

news

May 1, 2022 Competitive physics informed networks solve PDEs to accuracies previously out of reach for physics informed networks! Joint work with Qi Zeng and Spencer Bryngelson.
Apr 19, 2022 I’ll be giving a talk at the UPenn applied mathematics and computational science colloquium on Friday April 22nd.
Feb 19, 2022 I will be giving a (virtual) talk at the Stanford Applied Math Seminar on Feb 23.
Nov 28, 2021 Joint work with Jeffrey, Alistair, Yuanyuan, and Anima now on the arxiv. We extend competitive gradient descent to arbitrary numbers of agents, prove its local convergence, and show that it learns significantly more performant policies when used for multi-agent reinforcement learning.
Oct 26, 2021 Check out Jiawei’s new paper to learn how to initialize residual networks deterministically!

selected publications

  1. Sparse Cholesky factorization by Kullback-Leibler minimization
    Schäfer, Florian, Katzfuss, Matthias, and Owhadi, Houman
    SIAM Journal on Scientific Computing 2021
    arXiv
  2. Compression, inversion, and approximate PCA of dense kernel matrices at near-linear computational complexity
    Schäfer, Florian, Sullivan, T. J., and Owhadi, Houman
    Multiscale Model. Simul. 2021
    arXiv Blog
  3. Implicit competitive regularization in GANs
    Schäfer, Florian, Zheng, Hongkai, and Anandkumar, Anima
    In the 37th International Conference on Machine Learning (ICML 2020)
    arXiv Blog Slides
  4. Competitive Gradient Descent
    Schäfer, Florian, and Anandkumar, Anima
    In Advances in Neural Information Processing Systems 32
    arXiv HTML Blog