SlenderGNN: Accurate, Robust, and Interpretable GNN, and the Reasons for its Success

10/08/2022
by   Jaemin Yoo, et al.
0

Can we design a GNN that is accurate and interpretable at the same time? Could it also be robust to handle the case of homophily, heterophily, or even noisy edges without network effects? We propose SlenderGNN that has all desirable properties: (a) accurate, (b) robust, and (c) interpretable. For the reasons of its success, we had to dig deeper: The result is our GNNLin framework which highlights the fundamental differences among popular GNN models (e.g., feature combination, structural normalization, etc.) and thus reveals the reasons for the success of our SlenderGNN, as well as the reasons for occasional failures of other GNN variants. Thanks to our careful design, SlenderGNN passes all the 'sanity checks' we propose, and it achieves the highest overall accuracy on 9 real-world datasets of both homophily and heterophily graphs, when compared against 10 recent GNN models. Specifically, SlenderGNN exceeds the accuracy of linear GNNs and matches or exceeds the accuracy of nonlinear models with up to 64 times fewer parameters.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
05/17/2023

Edge Directionality Improves Learning on Heterophilic Graphs

Graph Neural Networks (GNNs) have become the de-facto standard tool for ...
research
07/17/2023

Examining the Effects of Degree Distribution and Homophily in Graph Learning Models

Despite a surge in interest in GNN development, homogeneity in benchmark...
research
03/01/2023

HyScale-GNN: A Scalable Hybrid GNN Training System on Single-Node Heterogeneous Architecture

Graph Neural Networks (GNNs) have shown success in many real-world appli...
research
01/17/2020

GraphLIME: Local Interpretable Model Explanations for Graph Neural Networks

Graph structured data has wide applicability in various domains such as ...
research
04/21/2023

GCNH: A Simple Method For Representation Learning On Heterophilous Graphs

Graph Neural Networks (GNNs) are well-suited for learning on homophilous...
research
09/26/2019

PairNorm: Tackling Oversmoothing in GNNs

The performance of graph neural nets (GNNs) is known to gradually decrea...

Please sign up or login with your details

Forgot password? Click here to reset