Projects per year
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.
Original language | English |
---|---|
Pages (from-to) | 788-803 |
Number of pages | 16 |
Journal | IEEE Transactions on Signal and Information Processing over Networks |
Volume | 10 |
Early online date | 2024 |
DOIs | |
Publication status | Published - 2024 |
MoE publication type | A1 Journal article-refereed |
Keywords
- Sensitivity analysis
- graph convolutional neural network
- graph shift operator
- structural perturbation
- sensitivity analysis
- Graph convolutional neural network
Fingerprint
Dive into the research topics of 'Graph Convolutional Neural Networks Sensitivity under Probabilistic Error Model'. Together they form a unique fingerprint.Projects
- 2 Active
-
AoF-ISAC: AI-assisted Waveform and Beamforming Design for Integrated Sensing and Communication
Ollila, E. (Principal investigator), Kuikka, J. (Project Member), Wang, X. (Project Member) & Wang, L. (Project Member)
01/01/2024 → 31/12/2026
Project: Academy of Finland: Other research funding
-
AI Based RAN: Towards Scalable and AI-Based Solutions for Beyond-5G Radio Access Networks
Vorobyov, S. (Principal investigator), Esfandiari, M. (Project Member), Hassas Irani, K. (Project Member) & Zhang, T. (Project Member)
01/01/2023 → 31/12/2025
Project: Academy of Finland: Other research funding