.png){kind=icon}
.png){kind=icon}
Return
Feature Selection of Network Data VIAl2,pRegularization
.png){kind=icon}
PRE
AI
DOI：10.1007/s12559-020-09763-z.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Feature selection is the process of selecting a subset of relevant features from the original feature set, and it plays an important role in handling high-dimensional data. In recent years, sparse learning-based feature selection approaches have been widely studied, and different regularizers have been proposed. Among these regularizers, it has been found thatl(2,p)(0 2,porm-based feature selection to deal with network data in an unsupervised scenario, and design an iterative algorithm using the framework of the alternating direction method of multipliers. In order to deal with the nonsmooth and non-Lipschitz continuous subproblem caused byl(2,p), we design a nonmonotone smoothing trust region algorithm and present its global convergence analysis. The extensive numerical experiments on real-world network datasets validate the effectiveness of the proposed model and algorithm.
Keywords:
Feature selection
Sparsity regularization
l(2p) (0 < p < 1) norm]
Smoothing trust region
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
4.3
Papers:
1.6K
Citations:
3.6K