Figure 9
From: Identifying vital nodes in complex networks by adjacency information entropy
Largest component experiments for spatial networks and classical networks. In (a), we can see that when the nodes in the network are removed according to URank, removing approximately 10% of the nodes reduced the largest component size of the network by approximately 90%. If the nodes are removed after being ranked by other metrics, more than 30% of the nodes need to be removed to achieve the same effect. In (b), the performance of our metric is slightly better than other metrics. In (c,f), URank and SH perform equally well. In (d,e), the performance of URank gradually gets better as the node removal ratio increases until the largest component size of the network reaches the minimum.
