mdeff/cnn_graph
A TensorFlow implementation of convolutional neural networks on arbitrary graphs using fast localized spectral filtering, published at NeurIPS 2016.
Not currently ranked — collecting fresh signals.
star history
This repository provides an efficient generalization of CNNs to non-Euclidean graph structures. The implementation uses TensorFlow and presents spectral filtering approaches for graph-structured data, with Jupyter notebooks for reproducing experiments on MNIST and other datasets. It builds on foundational graph neural network papers including spectral networks and deep convolutional networks on graph-structured data.
Frequently asked
- What is mdeff/cnn_graph?
- A TensorFlow implementation of convolutional neural networks on arbitrary graphs using fast localized spectral filtering, published at NeurIPS 2016.
- Is cnn_graph open source?
- Yes — mdeff/cnn_graph is open source, released under the MIT license.
- What language is cnn_graph written in?
- mdeff/cnn_graph is primarily written in Jupyter Notebook.
- How popular is cnn_graph?
- mdeff/cnn_graph has 1.4k stars on GitHub.
- Where can I find cnn_graph?
- mdeff/cnn_graph is on GitHub at https://github.com/mdeff/cnn_graph.