Low-Rank Preserving Projections

摘要: As one of the most popular dimensionality reduction techniques, locality preserving projections (LPP) has been widely used in computer vision and pattern recognition. However, practical applications, data is always corrupted by noises. For data, samples from same class may not be distributed nearest area, thus LPP lose its effectiveness. In this paper, it assumed that grossly noise matrix sparse. Based on these assumptions, we propose a novel method, named low-rank (LRPP) for image classification. LRPP learns weight projecting low-dimensional subspace. We use $\boldsymbol {L_{21}} $ norm as sparse constraint nuclear matrix. keeps global structure during procedure learned low rank can reduce disturbance noises data. learn robust subspace To verify performance classification, compare with state-of-the-art methods. The experimental results show effectiveness feasibility proposed method encouraging results.