Skip to content
This repository was archived by the owner on Feb 21, 2025. It is now read-only.

Files

Latest commit

23d0905 · Sep 26, 2022

History

History
This branch is 1690 commits behind dmlc/dgl:master.

geniepath

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
Aug 20, 2021
Sep 26, 2022
Sep 26, 2022
Sep 26, 2022

DGL Implementation of the GeniePath Paper

This DGL example implements the GNN model proposed in the paper GeniePath: Graph Neural Networks with Adaptive Receptive Paths.

Example implementor

This example was implemented by Kay Liu during his SDE intern work at the AWS Shanghai AI Lab.

Dependencies

  • Python 3.7.10
  • PyTorch 1.8.1
  • dgl 0.7.0
  • scikit-learn 0.23.2

Dataset

The datasets used for node classification are Pubmed citation network dataset (tranductive) and Protein-Protein Interaction dataset (inductive).

How to run

If want to train on Pubmed (transductive), run

python pubmed.py

If want to use a GPU, run

python pubmed.py --gpu 0

If want to train GeniePath-Lazy, run

python pubmed.py --lazy True

If want to train on PPI (inductive), run

python ppi.py

Performance

Dataset: Pubmed (ACC)

Method GeniePath
Paper 78.5%
DGL 73.0%

Dataset: PPI (micro-F1)

Method GeniePath GeniePath-lazy GeniePath-lazy-residual
Paper 0.9520 0.9790 0.9850
DGL 0.9729 0.9802 0.9798