Probabilistic and Regularized Graph Convolutional Networks
This is an incremental study for the graph learning community, as it builds on existing GCN methods without introducing new breakthroughs.
The paper investigates Graph Convolutional Networks (GCNs) by reproducing prior results and testing modifications like graph regularization and alternative convolution approaches, but fails to achieve improvements in performance on standard datasets.
This paper explores the recently proposed Graph Convolutional Network architecture proposed in (Kipf & Welling, 2016) The key points of their work is summarized and their results are reproduced. Graph regularization and alternative graph convolution approaches are explored. I find that explicit graph regularization was correctly rejected by (Kipf & Welling, 2016). I attempt to improve the performance of GCN by approximating a k-step transition matrix in place of the normalized graph laplacian, but I fail to find positive results. Nonetheless, the performance of several configurations of this GCN variation is shown for the Cora, Citeseer, and Pubmed datasets.