Robust Fisher Linear Discriminant Analysis with Generalized Correntropic Loss Function
Xuemei Qin, Shengqi Wang, Badong Chen, Kezhen Zhang · 2020
Fisher linear discriminant analysis (LDA) is widely used to solve classification problems. The classical LDA is developed based on the L2-norm, which is very sensitive to outliers. In this paper, we propose a new LDA with generalized correntropic loss function (GCLF), termed as LDA-GCLF, to improve the robustness of LDA. LDA-GCLF is realized by the alternating direction method of multipliers (ADMM) and the gradient descent method. Experiments results on synthetic and real-world datasets demonstrate the effectiveness of LDA-GCLF.