Supervised community detection with line graph neural networks

Zhengdao Chen, Joan Bruna Estrach, Lisha Li

Research output: Contribution to conferencePaper

Abstract

Community detection in graphs can be solved via spectral methods or posterior inference under certain probabilistic graphical models. Focusing on random graph families such as the stochastic block model, recent research has unified both approaches and identified both statistical and computational detection thresholds in terms of the signal-to-noise ratio. By recasting community detection as a node-wise classification problem on graphs, we can also study it from a learning perspective. We present a novel family of Graph Neural Networks (GNNs) for solving community detection problems in a supervised learning setting. We show that, in a data-driven manner and without access to the underlying generative models, they can match or even surpass the performance of the belief propagation algorithm on binary and multiclass stochastic block models, which is believed to reach the computational threshold in these cases. In particular, we propose to augment GNNs with the non-backtracking operator defined on the line graph of edge adjacencies. The GNNs are achieved good performance on real-world datasets. In addition, we perform the first analysis of the optimization landscape of using (linear) GNNs to solve community detection problems, demonstrating that under certain simplifications and assumptions, the loss value at any local minimum is close to the loss value at the global minimum/minima.

Original languageEnglish (US)
StatePublished - Jan 1 2019
Event7th International Conference on Learning Representations, ICLR 2019 - New Orleans, United States
Duration: May 6 2019May 9 2019

Conference

Conference7th International Conference on Learning Representations, ICLR 2019
CountryUnited States
CityNew Orleans
Period5/6/195/9/19

Fingerprint

neural network
Neural networks
community
Supervised learning
learning
performance
Values
Signal to noise ratio
Neural Networks
Graph

ASJC Scopus subject areas

  • Education
  • Computer Science Applications
  • Linguistics and Language
  • Language and Linguistics

Cite this

Chen, Z., Bruna Estrach, J., & Li, L. (2019). Supervised community detection with line graph neural networks. Paper presented at 7th International Conference on Learning Representations, ICLR 2019, New Orleans, United States.

Supervised community detection with line graph neural networks. / Chen, Zhengdao; Bruna Estrach, Joan; Li, Lisha.

2019. Paper presented at 7th International Conference on Learning Representations, ICLR 2019, New Orleans, United States.

Research output: Contribution to conferencePaper

Chen, Z, Bruna Estrach, J & Li, L 2019, 'Supervised community detection with line graph neural networks', Paper presented at 7th International Conference on Learning Representations, ICLR 2019, New Orleans, United States, 5/6/19 - 5/9/19.
Chen Z, Bruna Estrach J, Li L. Supervised community detection with line graph neural networks. 2019. Paper presented at 7th International Conference on Learning Representations, ICLR 2019, New Orleans, United States.
Chen, Zhengdao ; Bruna Estrach, Joan ; Li, Lisha. / Supervised community detection with line graph neural networks. Paper presented at 7th International Conference on Learning Representations, ICLR 2019, New Orleans, United States.
@conference{248a9e4d716f4226b19aae2a03dc4e15,
title = "Supervised community detection with line graph neural networks",
abstract = "Community detection in graphs can be solved via spectral methods or posterior inference under certain probabilistic graphical models. Focusing on random graph families such as the stochastic block model, recent research has unified both approaches and identified both statistical and computational detection thresholds in terms of the signal-to-noise ratio. By recasting community detection as a node-wise classification problem on graphs, we can also study it from a learning perspective. We present a novel family of Graph Neural Networks (GNNs) for solving community detection problems in a supervised learning setting. We show that, in a data-driven manner and without access to the underlying generative models, they can match or even surpass the performance of the belief propagation algorithm on binary and multiclass stochastic block models, which is believed to reach the computational threshold in these cases. In particular, we propose to augment GNNs with the non-backtracking operator defined on the line graph of edge adjacencies. The GNNs are achieved good performance on real-world datasets. In addition, we perform the first analysis of the optimization landscape of using (linear) GNNs to solve community detection problems, demonstrating that under certain simplifications and assumptions, the loss value at any local minimum is close to the loss value at the global minimum/minima.",
author = "Zhengdao Chen and {Bruna Estrach}, Joan and Lisha Li",
year = "2019",
month = "1",
day = "1",
language = "English (US)",
note = "7th International Conference on Learning Representations, ICLR 2019 ; Conference date: 06-05-2019 Through 09-05-2019",

}

TY - CONF

T1 - Supervised community detection with line graph neural networks

AU - Chen, Zhengdao

AU - Bruna Estrach, Joan

AU - Li, Lisha

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Community detection in graphs can be solved via spectral methods or posterior inference under certain probabilistic graphical models. Focusing on random graph families such as the stochastic block model, recent research has unified both approaches and identified both statistical and computational detection thresholds in terms of the signal-to-noise ratio. By recasting community detection as a node-wise classification problem on graphs, we can also study it from a learning perspective. We present a novel family of Graph Neural Networks (GNNs) for solving community detection problems in a supervised learning setting. We show that, in a data-driven manner and without access to the underlying generative models, they can match or even surpass the performance of the belief propagation algorithm on binary and multiclass stochastic block models, which is believed to reach the computational threshold in these cases. In particular, we propose to augment GNNs with the non-backtracking operator defined on the line graph of edge adjacencies. The GNNs are achieved good performance on real-world datasets. In addition, we perform the first analysis of the optimization landscape of using (linear) GNNs to solve community detection problems, demonstrating that under certain simplifications and assumptions, the loss value at any local minimum is close to the loss value at the global minimum/minima.

AB - Community detection in graphs can be solved via spectral methods or posterior inference under certain probabilistic graphical models. Focusing on random graph families such as the stochastic block model, recent research has unified both approaches and identified both statistical and computational detection thresholds in terms of the signal-to-noise ratio. By recasting community detection as a node-wise classification problem on graphs, we can also study it from a learning perspective. We present a novel family of Graph Neural Networks (GNNs) for solving community detection problems in a supervised learning setting. We show that, in a data-driven manner and without access to the underlying generative models, they can match or even surpass the performance of the belief propagation algorithm on binary and multiclass stochastic block models, which is believed to reach the computational threshold in these cases. In particular, we propose to augment GNNs with the non-backtracking operator defined on the line graph of edge adjacencies. The GNNs are achieved good performance on real-world datasets. In addition, we perform the first analysis of the optimization landscape of using (linear) GNNs to solve community detection problems, demonstrating that under certain simplifications and assumptions, the loss value at any local minimum is close to the loss value at the global minimum/minima.

UR - http://www.scopus.com/inward/record.url?scp=85071184449&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85071184449&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85071184449

ER -