https://github.com/datake/AdaGCN

Official Implementation of AdaGCN (ICLR 2021)
https://github.com/datake/AdaGCN

Last synced: about 2 months ago
JSON representation

Official Implementation of AdaGCN (ICLR 2021)

• Host: GitHub
• URL: https://github.com/datake/AdaGCN
• Owner: datake
• Created: 2021-01-23T06:45:07.000Z (over 3 years ago)
• Default Branch: main
• Last Pushed: 2022-01-03T17:06:54.000Z (over 2 years ago)
• Last Synced: 2024-07-01T22:16:29.703Z (3 months ago)
• Language: Python
• Homepage:
• Size: 8.21 MB
• Stars: 59
• Watchers: 4
• Forks: 12
• Open Issues: 2
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• awesome-gradient-boosting-papers - [Code

README

        
# Official Pytorch Implementation of "AdaGCN: Adaboosting Graph Convolutional Networks into Deep Models" (ICLR 2021)

Please refer to [openreview](https://openreview.net/forum?id=QkRbdiiEjM) (ICLR 2021) to look into the details of our paper.

![Alt text](https://github.com/datake/AdaGCN/raw/main/AdaGCN.png)

## Enviromment

```

python3.6  

cuda11.0  

torch1.7.1

```

## Run the code （Datasets: citeseer, cora_ml, pubmed and ms_academic）

#### Baseline: GCN

```

python main.py --trainsize 20 --dataset citeseer --niter 5 --nseed 20 --model GCN --dropout 0.5 --reg 5e-4

```

#### Baseline: APPNP or PPNP

```

python main.py --trainsize 20 --dataset citeseer --niter 5 --nseed 20 --model APPNP --dropout 0.5 --early 1 --patience 300 --max 500 --reg 5e-3

```

#### AdaGCN on Four datasets:

```

python main.py --trainsize 20 --dataset citeseer --niter 5 --nseed 20 --model AdaGCN --layers 15 --hid_AdaGCN 5000 --dropout 0.0 --weight_decay 1e-3 --early 1 --patience 300 --max 500 --reg 5e-3   

python main.py --trainsize 20 --dataset cora_ml --niter 5 --nseed 20 --model AdaGCN --layers 12 --hid_AdaGCN 5000 --dropout 0.0 --weight_decay 1e-4 --early 1 --patience 300 --max 500 --reg 5e-3   

python main.py --trainsize 20 --dataset pubmed --niter 5 --nseed 20 --model AdaGCN --layers 20 --hid_AdaGCN 5000 --dropout 0.2 --weight_decay 1e-4 --early 1 --patience 300 --max 500 --reg 5e-3   

python main.py --trainsize 20 --dataset ms_academic --niter 5 --nseed 20 --model AdaGCN --layers 5 --hid_AdaGCN 3000 --dropout 0.2 --weight_decay 1e-4 --early 1 --patience 300 --max 500 --reg 5e-3

```

**Results:** 

| Dataset  | Average Accuracy | Std |

| ------------- | ------------- | ------------- |

| Citeseer  | 76.68  | 0.20  |

| Cora-ML  | 85.97  | 0.20  |

| PubMed  | 79.95  | 0.21  |

| MS Academic  | 93.17  | 0.07  |

## Acknowledgement

Our code is directly adapted from PPNP paper **Predict then Propagate: Graph Neural Networks meet Personalized PageRank** (ICLR 2019) github: https://github.com/klicperajo/ppnp.

## Contact

Please refer to [email protected] in case you have any questions. 

## Cite

Please cite our paper if you use the model or this code in your own work:

```

@inproceedings{sun2020adagcn,

  title={AdaGCN: Adaboosting Graph Convolutional Networks into Deep Models},

  author={Sun, Ke and Zhu, Zhanxing and Lin, Zhouchen},

  booktitle={International Conference on Learning Representations},

  year={2020}

}

```