Get our free extension to see links to code for papers anywhere online!

Chrome logo  Add to Chrome

Firefox logo Add to Firefox

Fea2Fea: Exploring Structural Feature Correlations via Graph Neural Networks

Jun 24, 2021
Jiaqing Xie, Rex Ying

Share this with someone who'll enjoy it:

Structural features are important features in graph datasets. However, although there are some correlation analysis of features based on covariance, there is no relevant research on exploring structural feature correlation on graphs with graph neural network based models. In this paper, we introduce graph feature to feature (Fea2Fea) prediction pipelines in a low dimensional space to explore some preliminary results on structural feature correlation, which is based on graph neural network. The results show that there exists high correlation between some of the structural features. A redundant feature combination with initial node features, which is filtered by graph neural network has improved its classification accuracy in some graph datasets. We compare the difference between concatenation methods on connecting embeddings between features and show that the simplest is the best. We generalize on the synthetic geometric graphs and certify the results on prediction difficulty between two structural features.

* Submitted to ECML-PKDD 2021 Graph Embedding and Mining(GEM) workshop 

   Access Paper Source

Share this with someone who'll enjoy it: