|
|
|
| Improving Network Controllability: a Graph Convolutional Network Based Approach |
| LU Xinbiao, LIU Zecheng, CHEN Guiyun, YANG Tieliu, GAO Xing
|
| School of Artificial Intelligence and Automation, Hohai University, Nanjing 211100, China |
|
|
|
|
Abstract In order to improve network controllability, a network controllability improvement method based on graph convolutional neural network is proposed, in which a graph convolutional network is first trained to select appropriate nodes, and then edges are randomly added between these selected nodes. Numerical simulations are carried out on two representative complex network models. Compared with the traditional method in which edges are added randomly between all nodes, the proposed method greatly reduces the number of added edges, which is more efficient.
|
|
Received: 02 January 2024
Published: 10 December 2025
|
|
|
|
|
|
|
|
|
