Geometric renormalization reveals hidden network symmetries by scaling them down while retaining key features. Extended to weighted networks, in which link intensities matter, here the authors present empirical evidence and theory to justify selecting links with maximum weights across increasingly longer length scales to reduce resolution, enabling self-similar replicas and study of size-dependent phenomena.
- Muhua Zheng
- Guillermo García-Pérez
- M. Ángeles Serrano