A Graph Is More Than Its Nodes: Towards Structured Uncertainty-Aware Learning on Graphs

10/27/2022
by   Hans Hao-Hsun Hsu, et al.
0

Current graph neural networks (GNNs) that tackle node classification on graphs tend to only focus on nodewise scores and are solely evaluated by nodewise metrics. This limits uncertainty estimation on graphs since nodewise marginals do not fully characterize the joint distribution given the graph structure. In this work, we propose novel edgewise metrics, namely the edgewise expected calibration error (ECE) and the agree/disagree ECEs, which provide criteria for uncertainty estimation on graphs beyond the nodewise setting. Our experiments demonstrate that the proposed edgewise metrics can complement the nodewise results and yield additional insights. Moreover, we show that GNN models which consider the structured prediction problem on graphs tend to have better uncertainty estimations, which illustrates the benefit of going beyond the nodewise setting.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/08/2023

Graph Neural Networks Use Graphs When They Shouldn't

Predictions over graphs play a crucial role in various domains, includin...
research
10/12/2022

What Makes Graph Neural Networks Miscalibrated?

Given the importance of getting calibrated predictions and reliable unce...
research
02/14/2022

Graph Neural Networks for Graphs with Heterophily: A Survey

Recent years have witnessed fast developments of graph neural networks (...
research
02/09/2021

SLAPS: Self-Supervision Improves Structure Learning for Graph Neural Networks

Graph neural networks (GNNs) work well when the graph structure is provi...
research
10/26/2021

Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification

The interdependence between nodes in graphs is key to improve class pred...
research
10/04/2022

Generative Models and Learning Algorithms for Core-Periphery Structured Graphs

We consider core-periphery structured graphs, which are graphs with a gr...
research
06/28/2023

Graph Interpolation via Fast Fused-Gromovization

Graph data augmentation has proven to be effective in enhancing the gene...

Please sign up or login with your details

Forgot password? Click here to reset