2019/12/10 by Mohsen Ghassemi Parsa, Hadi Zare, Parsa, Mohsen Ghassemi +3 · 1 citation
Computer Science · #FOS: Computer and information sciences #Face and Expression Recognition #Machine Learning (cs.LG) #Machine Learning (stat.ML)
paper · pdf · doi:10.48550/arxiv.1912.05458
openalex publication_date 2019/12/10 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Feature selection methods have an important role on the readability of data\nand the reduction of complexity of learning algorithms. In recent years, a\nvariety of efforts are investigated on feature selection problems based on\nunsupervised viewpoint due to the laborious labeling task on large datasets. In\nthis paper, we propose a novel approach on unsupervised feature selection\ninitiated from the subspace clustering to preserve the similarities by\nrepresentation learning of low dimensional subspaces among the samples. A\nself-expressive model is employed to implicitly learn the cluster similarities\nin an adaptive manner. The proposed method not only maintains the sample\nsimilarities through subspace clustering, but it also captures the\ndiscriminative information based on a regularized regression model. In line\nwith the convergence analysis of the proposed method, the experimental results\non benchmark datasets demonstrate the effectiveness of our approach as compared\nwith the state of the art methods.\n