Topological structure identification method for complex networks with anomalous edges
SCIENTIA SINICA Informationis · 01.05.2026 · 56 (5) · 1101
DOI: 10.1360/ssi-2025-0451 ↗Kurz gefasst
Fachartikel in der Zeitschrift SCIENTIA SINICA Informationis, online erschienen am 17.04.2026. Verfasst von Yaozhong ZHENG und 2 weiteren Autor:innen. Die Arbeit nennt 41 Literaturangaben.
Abstract (Originalsprache)
The topology of complex networks serves as the backbone of secure control. Network topology identification based on nodal time-series data has become a research hotspot across interdisciplinary fields such as computer science, systems science, physics, and others. However, the presence of anomalous edges in the network influences the identification performance and even weakens the security and stability of network control. As a remedy, this paper proposes a strategy block-sparse Bayesian learning algorithm (S-Block-SBL) for identifying network topology with anomalous edges. With the aid of block-sparse Bayesian learning, the algorithm unifies identification results across all datasets, and a dataset purification strategy is proposed to substantially improve identification accuracy. It is proved that the loss function of the proposed S-Block-SBL is nonincreasing and convergent. Extensive experiments on benchmark networks demonstrate the algorithm’s effectiveness and superiority.
Zitieren
APA
ZHENG, Y., ZHANG, H., & HU, Y. (2026). Topological structure identification method for complex networks with anomalous edges. SCIENTIA SINICA Informationis, 56(5), 1101. https://doi.org/10.1360/ssi-2025-0451
Vancouver
ZHENG Y, ZHANG H, HU Y. Topological structure identification method for complex networks with anomalous edges. SCIENTIA SINICA Informationis. 2026;56(5):1101. doi:10.1360/ssi-2025-0451
BibTeX
@article{ZHENG2026,
author = {ZHENG, Yaozhong and ZHANG, Haitao and HU, Yuanyuan},
title = {Topological structure identification method for complex networks with anomalous edges},
journal = {SCIENTIA SINICA Informationis},
year = {2026},
volume = {56},
number = {5},
pages = {1101},
doi = {10.1360/ssi-2025-0451}
}
EUVIMED