This paper explores the use of machine learning to efficiently dismantle complex networks and predict their disintegration. A machine learning framework, GDM (Graph Dismantling with Machine learning), is developed and trained on smaller networks to identify topological patterns crucial for efficient dismantling of larger networks. The model outperforms existing heuristics and provides a quantitative early-warning signal for systemic risk, predicting the probability of system collapse.
Publisher
NATURE COMMUNICATIONS
Published On
Aug 31, 2021
Authors
Marco Grassia, Manlio De Domenico, Giuseppe Mangioni
Tags
machine learning
network dismantling
system collapse
topological patterns
predictive model
Related Publications
Explore these studies to deepen your understanding of the subject.
