Sparse Neighborhood Preserving Embedding via L2,1-Norm Minimization
Youpeng Zhou, Yulin Ding, Yifu Luo, Haoliang Ren · 2016
In this article, a novel unsupervised feature selection algorithm is proposed. We incorporate sparse neighborhood preserving embedding (SNPE) and L2,1-norm minimization into a joint framework for unsupervised feature selection. Neighborhood preserving embedding (NPE) is a subspace learning which aims at preserving the local neighborhood structure on the data manifold, while L2,1-norm regularization is performed to the transformation matrix to enable feature selection across all data samples. The convergence analysis of our method is provided and experimental results demonstrate the efficiency of our algorithm.