Return
Modular effective resistance-guided multi-scale graph representation learning
DOI:10.1016/j.patrec.2026.07.034.png)
Abstract
En 中文
• A modular effective resistance technique (MERT) for scalable graph learning.
• Fusion of fine- and coarse-grained structural information in graphs.
• Modular topology optimization for large-scale graph representation.
• State-of-the-art performance on small and large benchmark datasets.
Keywords:
Graph neural network
Multi-scale learning
Graph partition
Journal
IF:
3.3
Papers:
8.0K
Citations:
1.6W
Organization
Cited Papers
Collaborative graph neural networks for augmented graphs: A local-to-global perspective
PATTERN RECOGNITION
IF7.6
no more

