README.md

Learnable Graph Convolutional Network and Feature Fusion (LGCN-FF)

Introduction

• This is an implement of LGCN-FF with PyTorch, which was run on a machine with AMD R9-5900HX CPU, RTX 3080 16G GPU and 32G RAM. It has been accepted by Information Fusion.

Paper

Zhaoliang Chen, Lele Fu, Jie Yao, Wenzhong Guo, Claudia Plant and Shiping Wang. "Learnable Graph Convolutional Network and Feature Fusion for Multi-view Learning." Information Fusion (2023).

Requirements

• torch: 1.11.0 + cu115
• numpy: 1.20.1
• scipy: 1.6.2
• scikit-learn: 1.1.2

Running Examples

• For ALOI:

  python ./main.py --dataset-name ALOI --k 15 --epoch-num 180 
  

• For BBCnews

  python ./main.py --dataset-name BBCnews --k 20 --epoch-num 300
  

• For BBCsports

  python ./main.py --dataset-name BBCsports --k 14 --epoch-num 170
  

• For Wikipedia

  python ./main.py --dataset-name Wikipedia --k 20 --epoch-num 150
  

• For MSRC-v1

  python ./main.py --dataset-name MSRC-v1 --k 10 --epoch-num 130
  

• For MNIST

  python ./main.py --dataset-name MNIST --k 12 --epoch-num 320
  

Cite

@article{chen2023learnable,
  title = {Learnable Graph Convolutional Network and Feature Fusion for Multi-view Learning},
  author = {Chen, Zhaoliang and Fu, Lele and Yao, Jie and Guo, Wenzhong and Plant, Claudia and Wang, Shiping},
  journal = {Information Fusion},
  year = {2023},
  pages = {109-119},
  volume = {95},
}