Return
Randomized ising models for graph node representation
DOI:10.1016/j.neucom.2026.133195.png)
Abstract
En 中文
• Novel graph reservoir based on Ising models, amenable to physical implementations. • Theorems for convergence at zero temperature and in the presence of thermal noise. • Accuracy on node classification tasks compatible with GESNs and fully-trained GNNs. • Analysis of the robustness to thermal noise and impact of reservoir designs.
Keywords:
Ising models
graph reservoir
node classification
thermal noise
graph representation
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W

