摘要
Community structure is one of the most significant features in complex networks and community detection is a crucial method to analyze community structure. Existing representations in community detection have the characteristics of inflexibility and easily generate invalid solutions. To address the drawbacks, this paper proposed a multi-objective evolutionary algorithm based on mixed encoding (MOGAME). The algorithm combines the locus-based representation and labels-based representation, which can avoid generating invalid solution and improve the performance. Extensive experiments on both synthetic and real-word networks show that the proposed algorithm performs better than the existing algorithms with respect to accuracy and stability.
学科与关键词
关键词
Complex networkMulti-objective evolutionaryMixed encodingCommunityDetection
- A multi-objective adaptive evolutionary algorithm to extract communities in networksLi, Qi; Cao, Zehong*; Ding, Weiping*; Li, QingSwarm and Evolutionary Computation, 2020, 52: 100629.被引 30
- Evolutionary Computing Empowered Community Detection in Attributed NetworksGuo, Kun; Chen, Zhanhong; Yu, Zhiyong*; Chen, Kai; Guo, WenzhongIEEE Communications Magazine, 2024, 62(5): 22-26.
- Community Detection Based on Multiobjective Particle Swarm Optimization and Graph Attention Variational AutoencoderGuo, Kun; Chen, Zhanhong; Lin, Xu; Wu, Ling; Zhan, Zhi-Hui; Chen, Yuzhong; Guo, Wenzhong*IEEE Transactions on Big Data, 2023, 9(2): 569-583.被引 2
