.png){kind=icon}
.png){kind=icon}
Return
Semidefinite program-inspired continuous relaxation robust multi-view clustering for large-scale data
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2026.115646.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In the modern era, with diverse and massive information, a large number of unlabeled large-scale multi-view datasets have emerged. Multi-view Clustering (MVC) has the potential to handle this type of dataset. However, in existing research, MVC methods have insufficient consideration of robustness and scalability. To address these issues, we propose a novel Semi-definite Program-inspired Continuous Relaxation Robust Multi-view Clustering for Large-Scale Data (SPRMC-LS) whose complexity is linear with the sample number. This model integrates the anchor strategy into the kernel function to obtain the view-specific similarity matrix. An iteration-free SDP-inspired continuous relaxation approach makes the indicator matrix able to distinguish outliers. Finally, we use a simple binary classification function to fuse the results of each view. Experiments conducted on several datasets demonstrate the superiority of our SPRMC-LS.
Keywords:
Multi-view Clustering
Robustness
Scalability
Large-Scale Data
Semi-definite Programming
Journal
K
IF:
7.6
Papers:
1.2W
Citations:
4.5W