EUVIMEDEUVIMED entdecken
Crossref

ASWT-SGNN: Adaptive Spectral Wavelet Transform-Based Self-Supervised Graph Neural Network

Ruyue Liu, Rong Yin, Yong Liu, Weiping Wang

Proceedings of the AAAI Conference on Artificial Intelligence · 24.03.2024 · 38 (12) · 13990-13998

DOI: 10.1609/aaai.v38i12.29307 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 24.03.2024. Verfasst von Ruyue Liu und 3 weiteren Autor:innen. Die Arbeit wurde bisher 13-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Graph Comparative Learning (GCL) is a self-supervised method that combines the advantages of Graph Convolutional Networks (GCNs) and comparative learning, making it promising for learning node representations. However, the GCN encoders used in these methods rely on the Fourier transform to learn fixed graph representations, which is inherently limited by the uncertainty principle involving spatial and spectral localization trade-offs. To overcome the inflexibility of existing methods and the computationally expensive eigen-decomposition and dense matrix multiplication, this paper proposes an Adaptive Spectral Wavelet Transform-based Self-Supervised Graph Neural Network (ASWT-SGNN). The proposed method employs spectral adaptive polynomials to approximate the filter function and optimize the wavelet using contrast loss. This design enables the creation of local filters in both spectral and spatial domains, allowing flexible aggregation of neighborhood information at various scales and facilitating controlled transformation between local and global information. Compared to existing methods, the proposed approach reduces computational complexity and addresses the limitation of graph convolutional neural networks, which are constrained by graph size and lack flexible control over the neighborhood aspect. Extensive experiments on eight benchmark datasets demonstrate that ASWT-SGNN accurately approximates the filter function in high-density spectral regions, avoiding costly eigen-decomposition. Furthermore, ASWT-SGNN achieves comparable performance to state-of-the-art models in node classification tasks.

Zitieren

APA

Liu, R., Yin, R., Liu, Y., & Wang, W. (2024). ASWT-SGNN: Adaptive Spectral Wavelet Transform-Based Self-Supervised Graph Neural Network. Proceedings of the AAAI Conference on Artificial Intelligence, 38(12), 13990-13998. https://doi.org/10.1609/aaai.v38i12.29307

Vancouver

Liu R, Yin R, Liu Y, Wang W. ASWT-SGNN: Adaptive Spectral Wavelet Transform-Based Self-Supervised Graph Neural Network. Proceedings of the AAAI Conference on Artificial Intelligence. 2024;38(12):13990-13998. doi:10.1609/aaai.v38i12.29307

BibTeX
@article{Liu2024,
  author = {Liu, Ruyue and Yin, Rong and Liu, Yong and Wang, Weiping},
  title = {ASWT-SGNN: Adaptive Spectral Wavelet Transform-Based Self-Supervised Graph Neural Network},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2024},
  volume = {38},
  number = {12},
  pages = {13990--13998},
  doi = {10.1609/aaai.v38i12.29307}
}
Kostenlos starten →