PyTorch(Geometric) implementation of G^2GNN in "Imbalanced Graph Classification via Graph-of-Graph Neural Networks"

Yu Wang (Jack)

Last update: Nov 18, 2022

Related tags

Deep Learning G2GNN

Overview

This repository is an official PyTorch(Geometric) implementation of G^2GNN in "Imbalanced Graph Classification via Graph-of-Graph Neural Networks". The whole flowchart our model is visualized in the sequal.

We provide the motivation of graph of graphs via computing the homophily level of each constructed graph with different r-top similar graphs as direct neighborhoods. We can clearly see that edge homophily decreases as r increases because graphs with low topological similarity have higher chance to be selected as neighborhoods while they likely belong to different classes from corresponding center graphs. However, edge homophily even with r up to 5 is still [0.7, 0.8] and comparable to Citeseer dataset, indicating most edges in the constructed graphs connects graphs of the same class. Motivated by this observation, we perform GoG propagation on the generated kNN graph to aggregate neighboring graph information.

Requirements

PyTorch 1.10.0+cu113
PyTorch Geometric 2.0.2
Pytorch-scatter 2.0.9
NetworkX 2.6.3
Tqdm 4.62.3
Sklearn 1.0.1

Note that the version of PyTorch and PyTorch Geometric should be compatible and PyTorch Geometric is related to other packages, which requires to be installed beforehand. It is recommended to follow the installation instruction.

Run

To reproduce the performance comparison and the ablation study in the following Table and the Figure, run

bash run_{dataset}.sh

Code for "SRHEN: Stepwise-Refining Homography Estimation Network via Parsing Geometric Correspondences in Deep Latent Space"

SRHEN This is a better and simpler implementation for "SRHEN: Stepwise-Refining Homography Estimation Network via Parsing Geometric Correspondences in

1 Oct 28, 2022

Geometric Algebra package for JAX

JAXGA - JAX Geometric Algebra GitHub | Docs JAXGA is a Geometric Algebra package on top of JAX. It can handle high dimensional algebras by storing onl

36 Dec 22, 2022

Geometric Vector Perceptrons --- a rotation-equivariant GNN for learning from biomolecular structure

Geometric Vector Perceptron Implementation of equivariant GVP-GNNs as described in Learning from Protein Structure with Geometric Vector Perceptrons b

142 Dec 29, 2022

Computational modelling of ray propagation through optical elements using the principles of geometric optics (Ray Tracer)

Computational modelling of ray propagation through optical elements using the principles of geometric optics (Ray Tracer) Introduction By applying the

1 Jul 9, 2022

GeoTransformer - Geometric Transformer for Fast and Robust Point Cloud Registration

Comments

Overall_reweight and Batch_reweight are not running!!!

Hello Dear, Thank you for sharing your research paper source code here. I think code is only running with "reweight" instead. Otherwise we will get below error:

G2GNN/learn.py", line 192, in train loss.backward() UnboundLocalError: local variable 'loss' referenced before assignment

Please correct me if I am wrong.

Thank you so much.

opened by Mah-Zam 4

PyTorch(Geometric) implementation of G^2GNN in "Imbalanced Graph Classification via Graph-of-Graph Neural Networks"

Related tags

Overview

Requirements

Run

You might also like...

Code for "SRHEN: Stepwise-Refining Homography Estimation Network via Parsing Geometric Correspondences in Deep Latent Space"

Geometric Algebra package for JAX

Geometric Vector Perceptrons --- a rotation-equivariant GNN for learning from biomolecular structure

Computational modelling of ray propagation through optical elements using the principles of geometric optics (Ray Tracer)

GeoTransformer - Geometric Transformer for Fast and Robust Point Cloud Registration

Buffon’s needle: one of the oldest problems in geometric probability

Contains code for Deep Kernelized Dense Geometric Matching

ALBERT-pytorch-implementation - ALBERT pytorch implementation

An essential implementation of BYOL in PyTorch + PyTorch Lightning

Comments

Overall_reweight and Batch_reweight are not running!!!

Owner

Yu Wang (Jack)

A Temporal Extension Library for PyTorch Geometric

Geometric Deep Learning Extension Library for PyTorch

Pytorch Geometric Tutorials

Implementation of GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation (ICLR 2022).

Multi-Scale Geometric Consistency Guided Multi-View Stereo

Code for "FGR: Frustum-Aware Geometric Reasoning for Weakly Supervised 3D Vehicle Detection", ICRA 2021

Geometric Vector Perceptron --- a rotation-equivariant GNN for learning from biomolecular structure

Certifiable Outlier-Robust Geometric Perception

A geometric deep learning pipeline for predicting protein interface contacts.

3DMV jointly combines RGB color and geometric information to perform 3D semantic segmentation of RGB-D scans.