GraphMix: Regularized Training of Graph Neural Networks for Semi-Supervised Learning

25 Sep 2019Vikas VermaMeng QuAlex LambYoshua BengioJuho KannalaJian Tang

We present GraphMix, a regularization technique for Graph Neural Network based semi-supervised object classification, leveraging the recent advances in the regularization of classical deep neural networks. Specifically, we propose a unified approach in which we train a fully-connected network jointly with the graph neural network via parameter sharing, interpolation-based regularization, and self-predicted-targets... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK COMPARE
Node Classification CiteSeer with Public Split: fixed 20 nodes per class GraphMix Accuracy 74.52 # 2
Node Classification Coauthor CS GraphMix Accuracy 91.83 # 2
Node Classification Cora with Public Split: fixed 20 nodes per class GraphMix Accuracy 83.94 # 6
Node Classification PubMed with Public Split: fixed 20 nodes per class GCN(predicted-targets) Accuracy 82.42% # 1