Low-rank subspace learning based network community detection

作者：

Highlights：

• A novel community detection algorithm based on low-rank subspace learning is proposed.

• We design a low-rank decomposition strategy to find the lowest rank representation of all node vectors jointly in the geodesic space.

• The proposed method is robust to perturbations in networks and can better discriminate the community boundaries.

• Experiments prove the effectiveness of LRSCD compared with representative existing methods.

摘要

•A novel community detection algorithm based on low-rank subspace learning is proposed.•We design a low-rank decomposition strategy to find the lowest rank representation of all node vectors jointly in the geodesic space.•The proposed method is robust to perturbations in networks and can better discriminate the community boundaries.•Experiments prove the effectiveness of LRSCD compared with representative existing methods.

论文关键词：Complex network,Community detection,Low-rank representation,Subspace learning

论文评审过程：Received 24 August 2017, Revised 10 April 2018, Accepted 20 April 2018, Available online 24 April 2018, Version of Record 28 May 2018.

论文官网地址：https://doi.org/10.1016/j.knosys.2018.04.026