Examining indicators of complex network vulnerability across diverse attack scenarios

• The methodology we employed in this study involved the analysis of complex networks to understand their vulnerability and robustness against various attack scenarios.

• We investigated topological properties or indicators, such as shortest path length, modularity, efficiency, graph density, diameter, assortativity, and clustering coefficient.

• We examined how the growth or depletion of nodes and links based on specific attack criteria affected the robustness of the network, measured by the largest connected component (LCC) size and diameter.

• To quantify the contribution of indicators on LCC preservation, we used partial least squares discriminant analysis, considering possible correlations between indicators.

• Our analysis included 14 complex network datasets and 5 attack models, consistently finding that high modularity and disassortativity were prime indicators of vulnerability, in line with prior studies.

• We conclude by highlighting the implications of this research for designing computational models that achieve network robustness in the face of changing or unknown attack strategies.