Home
Name Modified Size InfoDownloads / Week
src 2025-06-04
train.py 2025-06-04 7.8 kB
README.md 2025-06-04 805 Bytes
Totals: 3 Items   8.6 kB 0

Graph Clustering via Similarity-Aware Embeddings (GCSAE)

Deep graph clustering via RBF-Optimized embeddings with self-supervised learning

Requirements

[!NOTE] Higher versions should be also compatible.

  • torch
  • torchvision
  • torchaudio
  • torch-scatter
  • torch-sparse
  • torch-cluster
  • munkres
  • kmeans-pytorch
  • Scipy
  • Scikit-learn
pip install -r requirements.txt

Model

framework

Reproduction

The same code can be used for Citeseer, Amazon-Photo and Amazon-Computers by changing the dataset name.

  • Cora !python train.py --runs 1 --dataset 'Computers' --hidden '512' --1_1 100 --l_2 --tau 0.5 --ns 0.5 --lr 0.0005 --epochs_sim 150 --epochs_cluster 150 --wd 1e-3 --alpha 0.9
Source: README.md, updated 2025-06-04