On Filter Size in Graph Convolutional Networks

Dinh V. Tran; Nicolò Navarin; Alessandro Sperduti

doi:10.1109/SSCI.2018.8628758

On Filter Size in Graph Convolutional Networks

Dinh V. Tran, Nicolò Navarin, Alessandro Sperduti

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

58 Scopus citations

Abstract

Recently, many researchers have been focusing on the definition of neural networks for graphs. The basic component for many of these approaches remains the graph convolution idea proposed almost a decade ago. In this paper, we extend this basic component, following an intuition derived from the well-known convolutional filters over multi-dimensional tensors. In particular, we derive a simple, efficient and effective way to introduce a hyper-parameter on graph convolutions that influences the filter size, i.e., its receptive field over the considered graph. We show with experimental results on real-world graph datasets that the proposed graph convolutional filter improves the predictive performance of Deep Graph Convolutional Networks.

Original language	English
Title of host publication	Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018
Editors	Suresh Sundaram
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1534-1541
Number of pages	8
ISBN (Electronic)	9781538692769
DOIs	https://doi.org/10.1109/SSCI.2018.8628758
State	Published - 2 Jul 2018
Externally published	Yes
Event	8th IEEE Symposium Series on Computational Intelligence, SSCI 2018 - Bangalore, India Duration: 18 Nov 2018 → 21 Nov 2018

Publication series

Name	Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018

Conference

Conference	8th IEEE Symposium Series on Computational Intelligence, SSCI 2018
Country/Territory	India
City	Bangalore
Period	18/11/18 → 21/11/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

Keywords

Convolutional neural networks for graphs
deep learning for graphs
graph convolution
graphs

ASJC Scopus subject areas

Artificial Intelligence
Theoretical Computer Science

Access to Document

10.1109/SSCI.2018.8628758

Cite this

Tran, D. V., Navarin, N., & Sperduti, A. (2018). On Filter Size in Graph Convolutional Networks. In S. Sundaram (Ed.), Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018 (pp. 1534-1541). Article 8628758 (Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SSCI.2018.8628758

Tran, Dinh V. ; Navarin, Nicolò ; Sperduti, Alessandro. / On Filter Size in Graph Convolutional Networks. Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018. editor / Suresh Sundaram. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1534-1541 (Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018).

@inproceedings{1ac6b6cac5174726b4af3657b33c9d93,

title = "On Filter Size in Graph Convolutional Networks",

abstract = "Recently, many researchers have been focusing on the definition of neural networks for graphs. The basic component for many of these approaches remains the graph convolution idea proposed almost a decade ago. In this paper, we extend this basic component, following an intuition derived from the well-known convolutional filters over multi-dimensional tensors. In particular, we derive a simple, efficient and effective way to introduce a hyper-parameter on graph convolutions that influences the filter size, i.e., its receptive field over the considered graph. We show with experimental results on real-world graph datasets that the proposed graph convolutional filter improves the predictive performance of Deep Graph Convolutional Networks.",

keywords = "Convolutional neural networks for graphs, deep learning for graphs, graph convolution, graphs",

author = "Tran, \{Dinh V.\} and Nicol{\`o} Navarin and Alessandro Sperduti",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 8th IEEE Symposium Series on Computational Intelligence, SSCI 2018 ; Conference date: 18-11-2018 Through 21-11-2018",

year = "2018",

month = jul,

day = "2",

doi = "10.1109/SSCI.2018.8628758",

language = "English",

series = "Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1534--1541",

editor = "Suresh Sundaram",

booktitle = "Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018",

address = "United States",

}

Tran, DV, Navarin, N & Sperduti, A 2018, On Filter Size in Graph Convolutional Networks. in S Sundaram (ed.), Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018., 8628758, Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018, Institute of Electrical and Electronics Engineers Inc., pp. 1534-1541, 8th IEEE Symposium Series on Computational Intelligence, SSCI 2018, Bangalore, India, 18/11/18. https://doi.org/10.1109/SSCI.2018.8628758

On Filter Size in Graph Convolutional Networks. / Tran, Dinh V.; Navarin, Nicolò; Sperduti, Alessandro.
Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018. ed. / Suresh Sundaram. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1534-1541 8628758 (Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - On Filter Size in Graph Convolutional Networks

AU - Tran, Dinh V.

AU - Navarin, Nicolò

AU - Sperduti, Alessandro

PY - 2018/7/2

Y1 - 2018/7/2

N2 - Recently, many researchers have been focusing on the definition of neural networks for graphs. The basic component for many of these approaches remains the graph convolution idea proposed almost a decade ago. In this paper, we extend this basic component, following an intuition derived from the well-known convolutional filters over multi-dimensional tensors. In particular, we derive a simple, efficient and effective way to introduce a hyper-parameter on graph convolutions that influences the filter size, i.e., its receptive field over the considered graph. We show with experimental results on real-world graph datasets that the proposed graph convolutional filter improves the predictive performance of Deep Graph Convolutional Networks.

AB - Recently, many researchers have been focusing on the definition of neural networks for graphs. The basic component for many of these approaches remains the graph convolution idea proposed almost a decade ago. In this paper, we extend this basic component, following an intuition derived from the well-known convolutional filters over multi-dimensional tensors. In particular, we derive a simple, efficient and effective way to introduce a hyper-parameter on graph convolutions that influences the filter size, i.e., its receptive field over the considered graph. We show with experimental results on real-world graph datasets that the proposed graph convolutional filter improves the predictive performance of Deep Graph Convolutional Networks.

KW - Convolutional neural networks for graphs

KW - deep learning for graphs

KW - graph convolution

KW - graphs

UR - https://www.scopus.com/pages/publications/85062786741

U2 - 10.1109/SSCI.2018.8628758

DO - 10.1109/SSCI.2018.8628758

M3 - Conference contribution

AN - SCOPUS:85062786741

T3 - Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018

SP - 1534

EP - 1541

BT - Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018

A2 - Sundaram, Suresh

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 8th IEEE Symposium Series on Computational Intelligence, SSCI 2018

Y2 - 18 November 2018 through 21 November 2018

ER -

Tran DV, Navarin N, Sperduti A. On Filter Size in Graph Convolutional Networks. In Sundaram S, editor, Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1534-1541. 8628758. (Proceedings of the 2018 IEEE Symposium Series on Computational Intelligence, SSCI 2018). doi: 10.1109/SSCI.2018.8628758

On Filter Size in Graph Convolutional Networks

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this