Skip to content

Latest commit

 

History

History
24 lines (16 loc) · 756 Bytes

README.md

File metadata and controls

24 lines (16 loc) · 756 Bytes

magnesium

Multi Graph Convolutional Neural Networks (MGCNN) applied to Matrix Completion.

Purpose

The code contained in this repository represents a TensorFlow implementation of the Recurrent Muli-Graph Convolutional Neural Network depicted in:

Geometric Matrix Completion with Recurrent Multi-Graph Neural Networks (in Proc. NIPS, 2017)
Federico Monti, Michael M. Bronstein, Xavier Bresson
https://arxiv.org/abs/1704.06803

License

GNU General Public License v3.0 by @fmonti (Frederico Monti)

Modifications

@dtsbourg (Dylan Bourgeois)

This code is an attempt to package the code presented in https://github.com/fmonti/mgcnn for the Netflix challenge.