divelab / lgcn

Large-Scale Learnable Graph Convolutional Networks(LGCN)

Created by Hongyang Gao, Zhengyang Wang and Shuiwang Ji at Texas A&M University.

Accepted by KDD18.

Introduction

Large-Scale Learnable Graph Convolutional Networks provide an efficient way (LGCL and LGCN) for learnable graph convolution.

Detailed information about LGCL and LGCN is provided in (https://dl.acm.org/citation.cfm?id=3219947).

Methods

In this work, we propose the learnable graph convolution layer (LGCL). Based on LGCL. We propose the learnable graph convolutional networks.

Learnable Graph Convolution Layer

Learnable graph Convolutional Networks

Batch Training

Citation

If using this code, please cite our paper.

@inproceedings{gao2018large,
  title={Large-Scale Learnable Graph Convolutional Networks},
  author={Gao, Hongyang and Wang, Zhengyang and Ji, Shuiwang},
  booktitle={Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining},
  pages={1416--1424},
  year={2018},
  organization={ACM}
}

Start training

After configure the network, we can start to train. Run

python main.py

The training results on Cora dataset will be displayed.

Results

Models	Cora	Citeseer	Pubmed
DeepWalk	67.2%	43.2%	65.3%
Planetoid	75.7%	64.7%	77.2%
Chebyshev	81.2%	69.8%	74.4%
GCN	81.5%	70.3%	79.0%
LGCN	83.3 ± 0.5%	73.0 ± 0.6%	79.5 ± 0.2%