.png){kind=icon}
.png){kind=icon}
Return
Quantum entropy structural encoding for graph neural networks
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2025.114580.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Introduces QESE, a novel structural encoding method that enhances GNN expressive power beyond 3-WL using quantum information theory. • Demonstrates QESE’s ability to capture node subgraphs, global node positions, and graph structures, outperforming existing SE methods in graph learning tasks. • Proposes an approximation method QESE * which holds similar expressive power. Designs a plug-and-play approach to inject QESE into any existing GNNs.
Journal
K
IF:
7.6
Papers:
1.2W
Citations:
4.5W