Robust Sparse Linear Discriminant Analysis

2018/01/29 by Jie Wen, Xiaozhao Fang, Jinrong Cui +4 · 4 citations
Computer Science · Engineering · Social Sciences · #Face and Expression Recognition #Remote-Sensing Image Classification #Advanced Computing and Algorithms

paper · doi:10.1109/tcsvt.2018.2799214

Abstract

Linear discriminant analysis (LDA) is a very popular supervised feature extraction method and has been extended to different variants. However, classical LDA has the following problems: 1) The obtained discriminant projection does not have good interpretability for features; 2) LDA is sensitive to noise; and 3) LDA is sensitive to the selection of number of projection directions. In this paper, a novel feature extraction method called robust sparse linear discriminant analysis (RSLDA) is proposed to solve the above problems. Specifically, RSLDA adaptively selects the most discriminative features for discriminant analysis by introducing thel_2,1norm. An orthogonal matrix and a sparse matrix are also simultaneously introduced to guarantee that the extracted features can hold the main energy of the original data and enhance the robustness to noise, and thus RSLDA has the potential to perform better than other discriminant methods. Extensive experiments on six databases demonstrate that the proposed method achieves the competitive performance compared with other state-of-the-art feature extraction methods. Moreover, the proposed method is robust to the noisy data.

Citations

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Sparse Discriminant Analysis
• Robust principal component analysis?
• ImageNet classification with deep convolutional neural networks
• Compressed sensing

Cited by

• A supervised discriminant data representation: application to pattern classification
• Incomplete Multi-view Clustering via Graph Regularized Matrix Factorization
• Visual Analytics Using Tensor Unified Linear Comparative Analysis
• Self-Weighted Robust LDA for Multiclass Classification with Edge Classes

Related

• Robust Linear Discriminant Analysis Using Ratio Minimization of L1,2-Norms
• A supervised discriminant data representation: application to pattern classification
• An Efficient Approach to Sparse Linear Discriminant Analysis
• Spatial Sign based Direct Sparse Linear Discriminant Analysis for High Dimensional Data
• A Direct Estimation Approach to Sparse Linear Discriminant Analysis