SRAGL-AWCL: A two-step multi-view clustering via sparse representation and adaptive weighted cooperative learning

Highlights：

• We propose a novel multi-view spectral clustering which combines sparse representation by adaptive graph learning with adaptive weighting cooperative learning.

• In order to solve the problem that non-negative matrix factorization is sensitive to input data, we propose a new sparse decomposition model by removing non-negative constraint of the base matrix.

• To extract unique information of each view, the adaptive weighting method is used to learn a global matrix for fusion of different views. Then, we further optimize this global matrix by the symmetry operation.

• The proposed algorithms have obtained very good experimental results in several well-known single-view and multi-view datasets.

摘要

•We propose a novel multi-view spectral clustering which combines sparse representation by adaptive graph learning with adaptive weighting cooperative learning.•In order to solve the problem that non-negative matrix factorization is sensitive to input data, we propose a new sparse decomposition model by removing non-negative constraint of the base matrix.•To extract unique information of each view, the adaptive weighting method is used to learn a global matrix for fusion of different views. Then, we further optimize this global matrix by the symmetry operation.•The proposed algorithms have obtained very good experimental results in several well-known single-view and multi-view datasets.

论文关键词：Multi-view clustering,Sparse representation (sr),Adaptive graph learning (agl),Adaptive weighted cooperative learning (awcl),Global Optimized Matrix

论文评审过程：Received 31 October 2019, Revised 13 February 2021, Accepted 5 April 2021, Available online 14 April 2021, Version of Record 23 April 2021.