Percolation on complex networks studies how the giant cluster of a network disappears under removing nodes or links. There are various strategies to select removing node, for instance, random failure, high-degree attack, PageRank(PR), k-core and Collective Influence(CI)-based attack, etc.\\ Here...
Percolation on complex networks studies how the giant cluster of a network disappears under removing nodes or links. There are various strategies to select removing node, for instance, random failure, high-degree attack, PageRank(PR), k-core and Collective Influence(CI)-based attack, etc.\\ Here, we study among other things the so-called CI-based attack, or CI percolation, on random networks, which was introduced as a heuristic method for optimal percolation [Morone and Makse, Nature 524, 6568 (2015)]. The CI percolation is based on the value so-called ‘CI’, defined as CI$_{\ell}\left(i\right)=\left(k_{i}-1\right)\sum \limits_{j \in \partial B \left( i , \ell \right)}^{} \left(k_{j}-1\right)$, which incorporates both the degree of the node itself and its neighboring nodes at distance $\ell$. In CI percolation, the node having the largest value CI is called ‘top influencer’ and gets eliminated successively. As a result, the giant cluster collapses more quickly and suddenly, so that the percolation threshold becomes smaller than most of the other previously-studied strategies. We numerically investigate the critical behavior at the CI percolation transition. We estimate various critical exponent for $\ell$=1 and discuss the physical meaning of the obtained results.
Percolation on complex networks studies how the giant cluster of a network disappears under removing nodes or links. There are various strategies to select removing node, for instance, random failure, high-degree attack, PageRank(PR), k-core and Collective Influence(CI)-based attack, etc.\\ Here, we study among other things the so-called CI-based attack, or CI percolation, on random networks, which was introduced as a heuristic method for optimal percolation [Morone and Makse, Nature 524, 6568 (2015)]. The CI percolation is based on the value so-called ‘CI’, defined as CI$_{\ell}\left(i\right)=\left(k_{i}-1\right)\sum \limits_{j \in \partial B \left( i , \ell \right)}^{} \left(k_{j}-1\right)$, which incorporates both the degree of the node itself and its neighboring nodes at distance $\ell$. In CI percolation, the node having the largest value CI is called ‘top influencer’ and gets eliminated successively. As a result, the giant cluster collapses more quickly and suddenly, so that the percolation threshold becomes smaller than most of the other previously-studied strategies. We numerically investigate the critical behavior at the CI percolation transition. We estimate various critical exponent for $\ell$=1 and discuss the physical meaning of the obtained results.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.