ゲノム情報科学研究教育機構  アブストラクト
Date 2:00pm January 31, 2023
Speaker Yang Lou
Osaka University
Title Evaluating the Network Robustness: A Convolutional Neural Network Approach
Abstract
Network robustness is critical for various industrial and social networks against malicious attacks, which has various meanings in different research contexts and here it refers to the ability of a network to sustain its functionality, for example connectivity, controllability and communication ability, when a fraction of the network fail to work due to attacks. The rapid development of complex networks research has stimulated special interest and great concern about the network robustness, which is essential for further analyzing and optimizing network structures towards engineering applications. Network robustness is quantified by a sequence of values that record the remaining network functionality of the network after a sequence of node- or edge-removal attacks. Traditionally, robustness is determined by attack simulations, which are computationally very time-consuming or even practically infeasible for large-scale networks. In this talk, an approach for network robustness estimation is presented, which is developed based on the convolutional neural networks (CNN). This approach has three different schemes from a developmental point of view, including the initial straightforward scheme, the learning feature-assisted scheme, and the pyramid pooling-assisted scheme. Extensive experimental studies on various synthetic and real-world networks demonstrate that: 1) the CNN-based approach is able to estimate the network robustness with low errors; 2) the runtime of CNN-based estimation is significantly lower than that of the attack simulations; and 3) it provides a good indicator for robustness, better than the classical spectral measures.

「セミナー」に戻る      
 ホーム