计算机应用与软件2023，Vol.40Issue(12)：299-304,6.DOI:10.3969/j.issn.1000-386x.2023.12.044
两级联合融合的多视图子空间聚类改进算法
IMPROVED MULTI-VIEW SUBSPACE CLUSTERING ALGORITHM BASED ON TWO-LEVEL JOINT FUSION
刘浩翰 1杜嘉欣 1李建伏1
作者信息
- 1. 中国民航大学计算机科学与技术学院 天津 300300
- 折叠
摘要
Abstract
The multi-view deep subspace clustering networks(MvDSCN)algorithm does not make full use of the complementary information of the multi-view,and the clustering results are directly obtained by performing a clustering,and only the data-level multi-view information fusion is considered,thereby reducing the clustering performance.To solve these problems,we proposed an improved MvDSCN algorithm based on two-level joint fusion(TJ-MvDSCN).We not only paid attention to the common information of the multi-view,but also the complementary information of the multi-view.We added the multi-view information fusion of the partition level to form a two-level fusion structure with the existing data-level information.We added the clustering loss,and based on the iterative optimization strategy,a multi-view clustering framework that could jointly learn feature representation and clustering assignment was constructed.The experimental results show that the performance of the algorithm is better than existing algorithms.关键词
多视图聚类/子空间聚类/两级融合/联合学习Key words
Multi-view clustering/Subspace clustering/Two-level fusion/Joint learning分类
信息技术与安全科学引用本文复制引用
刘浩翰,杜嘉欣,李建伏..两级联合融合的多视图子空间聚类改进算法[J].计算机应用与软件,2023,40(12):299-304,6.
