.png){kind=icon}
Return
Adaptive sparse graph learning for multi-view spectral clustering
.png){kind=icon}
PRE
AI
DOI：10.1007/s10489-022-04267-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In recent years, graph-based multi-view spectral clustering methods have achieved remarkable progress. However, most of which are generally deficient in the following two aspects. First, ignoring the different importance of multiple views, low-quality views in the multi-view data may seriously affect the clustering performance. Second, for constructed graphs, noise and outliers are difficult to effectively filter out. For the purpose of overcoming the above two deficiencies, this paper proposes a novel adaptive sparse graph learning for multi-view spectral clustering (ASGL) method. Specifically, the adaptive neighbor graph learning method is adopted to construct the similarity matrices of all views, which improves the robustness to noise and outliers. By adaptively assigning the weight of each view, the complementary information between the views is combined to more accurately describe the essential category attributes between the sample data. An effective algorithm for solving the optimization problem of ASGL model is proposed. Compared to several state-of-the-art algorithms, extensive experiments on several benchmark datasets verify good clustering performance of ASGL.
Keywords:
Multi-view spectral clustering
Adaptive sparse graph
Sparse representation
Consensus similarity matrix
Journal
IF:
3.5
Papers:
7.6K
Citations:
1.7W
Organization
Cited Papers
Multi-view subspace clustering with inter-cluster consistency and intra-cluster diversity among views.png)
APPLIED INTELLIGENCE
APPLIED INTELLIGENCEIF3.5
PREAI
ShareSaveEnhancing multi-view clustering through common subspace integration by considering both global similarities and local structuresNEUROCOMPUTING
NEUROCOMPUTINGIF6.5
PREAI
ShareSave
View more