ICML 2019oral111 citations
Stochastic Blockmodels meet Graph Neural Networks
Nikhil Mehta, Lawrence Carin Duke, Piyush Rai
Abstract
Stochastic blockmodels (SBM) and their variants, $e.g.$, mixed-membership and overlapping stochastic blockmodels, are latent variable based generative models for graphs. They have proven to be successful for various tasks, such as discovering the community structure and link prediction on graph-structured data. Recently, graph neural networks, $e.g.$, graph convolutional networks, have also emerged as a promising approach to learn powerful representations (embeddings) for the nodes in the graph, by exploiting graph properties such as locality and invariance. In this work, we unify these two directions by developing a
BibTeX
@InProceedings{pmlr-v97-mehta19a,
title = {Stochastic Blockmodels meet Graph Neural Networks},
author = {Mehta, Nikhil and Duke, Lawrence Carin and Rai, Piyush},
booktitle = {Proceedings of the 36th International Conference on Machine Learning},
pages = {4466--4474},
year = {2019},
editor = {Chaudhuri, Kamalika and Salakhutdinov, Ruslan},
volume = {97},
series = {Proceedings of Machine Learning Research},
month = {09--15 Jun},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v97/mehta19a/mehta19a.pdf},
url = {https://proceedings.mlr.press/v97/mehta19a.html},
abstract = {Stochastic blockmodels (SBM) and their variants, $e.g.$, mixed-membership and overlapping stochastic blockmodels, are latent variable based generative models for graphs. They have proven to be successful for various tasks, such as discovering the community structure and link prediction on graph-structured data. Recently, graph neural networks, $e.g.$, graph convolutional networks, have also emerged as a promising approach to learn powerful representations (embeddings) for the nodes in the graph, by exploiting graph properties such as locality and invariance. In this work, we unify these two directions by developing a