Graph Convolutional Neural Networks Sensitivity under Probabilistic Error Model

Loading...
Thumbnail Image

Access rights

openAccess
publishedVersion

URL

Journal Title

Journal ISSN

Volume Title

A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä

Date

2024

Department

Department of Information and Communications Engineering

Major/Subject

Mcode

Degree programme

Language

en

Pages

16

Series

IEEE Transactions on Signal and Information Processing over Networks, Volume 10, pp. 788-803

Abstract

Graph Neural Networks (GNNs), particularly Graph Convolutional Neural Networks (GCNNs), have emerged as pivotal instruments in machine learning and signal processing for processing graph-structured data. This paper proposes an analysis framework to investigate the sensitivity of GCNNs to probabilistic graph perturbations, directly impacting the graph shift operator (GSO). Our study establishes tight expected GSO error bounds, which are explicitly linked to the error model parameters, and reveals a linear relationship between GSO perturbations and the resulting output differences at each layer of GCNNs. This linearity demonstrates that a single-layer GCNN maintains stability under graph edge perturbations, provided that the GSO errors remain bounded, regardless of the perturbation scale. For multilayer GCNNs, the dependency of system's output difference on GSO perturbations is shown to be a recursion of linearity. Finally, we exemplify the framework with the Graph Isomorphism Network (GIN) and Simple Graph Convolution Network (SGCN). Experiments validate our theoretical derivations and the effectiveness of our approach.

Description

Publisher Copyright: © 2015 IEEE.

Keywords

Sensitivity analysis, graph convolutional neural network, graph shift operator, structural perturbation, sensitivity analysis, Graph convolutional neural network

Other note

Citation

Wang, X, Ollila, E & Vorobyov, S A 2024, ' Graph Convolutional Neural Networks Sensitivity under Probabilistic Error Model ', IEEE Transactions on Signal and Information Processing over Networks, vol. 10, pp. 788-803 . https://doi.org/10.1109/TSIPN.2024.3485532