Correlation analysis of combined layers in multiplex networks based on entropy

Area of Science:

• Network Science
• Information Theory
• Data Analysis

Background:

• Multiplex networks exhibit complex structural features due to inter-layer interactions.
• Link overlaps between layers are a prominent feature influencing network behavior.
• Quantifying the relationship between structural interactions and network behavior is crucial.

Purpose of the Study:

• To propose a novel structure entropy metric for multiplex networks.
• To evaluate the correlation between network layers using this new metric.
• To demonstrate the metric's ability to predict network behavior.

Main Methods:

• A new structure entropy is developed by incorporating overlapping links across multiplex network layers.
• The proposed structure entropy is used to quantify inter-layer correlations.
• The method's efficacy is tested on diverse real-world multiplex network datasets.

Main Results:

• The proposed structure entropy effectively captures correlations between multiplex network layers.
• The calculated structure entropy aligns with observed network behaviors.
• Validation across four distinct real-world multiplex networks confirms the method's applicability.

Conclusions:

• The novel structure entropy provides a robust measure for analyzing multiplex network inter-layer relationships.
• This metric offers valuable insights into how structural interactions influence overall network dynamics.
• The method is broadly applicable to various types of multiplex networks.