A Multiagent Memetic Optimization Algorithm Based on Temporal Asymptotic Surprise in Complex Networks to Reveal the...

Area of Science:

• Complex network analysis
• Graph theory
• Computational intelligence

Background:

• Complex networks are prevalent in real-world applications, necessitating community structure analysis.
• Dynamic networks evolve over time, posing challenges for traditional static community detection methods.
• Tracking and detecting communities in dynamic graphs is crucial for understanding network evolution.

Purpose of the Study:

• To propose a novel algorithm for detecting dynamic communities in complex networks.
• To address the challenge of community evolution in real-world dynamic networks.
• To develop a method that does not require prior knowledge of the number of communities.

Main Methods:

• A multiagent optimization memetic algorithm, termed DYNMAMA (dynamic multiagent memetic algorithm), is proposed.
• The algorithm operates on dynamic graph data, processing temporal information.
• Temporal asymptotic surprise is employed as the evaluation function for community detection.

Main Results:

• DYNMAMA effectively detects dynamic communities in both real-world and synthetic networks.
• The algorithm demonstrates superior convergence and optimality compared to existing approaches.
• It successfully identifies community structures without needing to pre-define the number of communities.

Conclusions:

• DYNMAMA offers an effective solution for dynamic community detection in complex networks.
• The proposed method is robust and adaptable to evolving network structures.
• It advances the field by providing a more convergent and optimal approach for dynamic network analysis.