Become a sponsor to Michaël Defferrard
I'm a scientist and a public servant. All my work is open and public-first. If my work enables yours, please consider sharing benefits!
My work mainly consists in:
- publishing findings with supporting experimental code, and
- consolidating tried and tested findings in maintained high-quality Python packages.
As I believe knowledge to be a public good, I share educational material, mentor projects, and disseminate knowledge by giving talks and organizing events. Twitter is my public discourse's platform of choice. Do reach out!
In the larger OSS world, I try to fix the problems I see. Mostly in the (python) scientific stack, but occasionally from distro maintenance to kernel drivers, passing through the software stack I depend on. Not that it matters, but I use Arch Linux and neovim. 😉
My ability to help you is not bound to financial contributions. Thanks for your support! ❤️
1 sponsor has funded mdeff’s work.
Featured work
-
mdeff/cnn_graph
Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering
Jupyter Notebook 1,320 -
epfl-lts2/pygsp
Graph Signal Processing in Python
Python 471 -
epfl-lts2/pyunlocbox
A Python convex optimization package using proximal splitting methods
Python 110 -
deepsphere/deepsphere-cosmo-tf1
A spherical convolutional neural network for cosmology (TFv1).
Jupyter Notebook 130 -
mdeff/ntds_2019
Material for the EPFL master course "A Network Tour of Data Science", edition 2019.
Jupyter Notebook 66