Finding influential nodes in complex networks based on Kullback-Leibler model within the neighborhood

Area of Science:

• Network Security
• Complex Network Analysis
• Machine Learning

Background:

• Federated learning involves extensive information exchange across distributed devices.
• Identifying influential nodes is crucial for optimizing reliability in federated learning systems.

Purpose of the Study:

• To propose a novel method for identifying influential nodes in complex networks.
• To enhance the reliability and efficiency of federated learning systems.

Main Methods:

• The Kullback-Leibler divergence within the neighborhood (KLN) algorithm simulates node failure.
• KLN quantifies information entropy loss using KL divergence and network attributes to assess node importance.

Main Results:

• KLN demonstrates superior accuracy and applicability across various network scales compared to 11 other algorithms.
• Experimental validation using the SIR model and a real-world epidemic network confirms KLN's effectiveness.

Conclusions:

• The KLN algorithm provides an effective approach for evaluating node importance in complex networks.
• Findings support the development of targeted management and control strategies for network security and epidemic prevention.