Identifying the influential nodes to fragment a network is of significant importance in hindering the spread of epidemics and maximizing the influence of advertisements. Here, we address the problem by minimizing the largest eigenvalue of the non-backtracking matrices of networks, based on which a novel method is proposed to identify the optimal influential nodes. Interestingly, the proposed method could find some particular nodes to fragment complex networks. Experiments in real networks demonstrate the effectiveness of the proposed method.

中文翻译：

---

基于非回溯矩阵的复杂网络分段

识别有影响力的节点以使网络分裂对于阻止流行病的传播和最大程度地发挥广告的影响具有重要意义。在这里，我们通过最小化网络的非回溯矩阵的最大特征值来解决该问题，在此基础上，提出了一种识别最佳影响节点的新方法。有趣的是，所提出的方法可以找到一些特定的节点来分割复杂的网络。真实网络中的实验证明了该方法的有效性。