2024-11-282024-11-282020Martín-Clemente, R. y Zarzoso, V. (2020). LDA via L1-PCA of Whitened Data. IEEE Transactions on Signal Processing, 68, 225 / 8911464-240. https://doi.org/10.1109/TSP.2019.2955860.1053-587X1941-0476https://hdl.handle.net/11441/165101This work is licensed under a Creative Commons Attribution 4.0 LicensePrincipal component analysis (PCA) and Fisher's linear discriminant analysis (LDA) are widespread techniques in data analysis and pattern recognition. Recently, the L1-norm has been proposed as an alternative criterion to classical L2-norm in PCA, drawing considerable research interest on account of its increased robustness to outliers. The present work proves that, combined with a whitening preprocessing step, L1-PCA can perform LDA in an unsupervised manner, i.e., sparing the need for labelled data. Rigorous proof is given in the case of data drawn from a mixture of Gaussians. A number of numerical experiments on synthetic as well as real data confirm the theoretical findings.application/pdf16 p.engAttribution 4.0 Internationalhttp://creativecommons.org/licenses/by/4.0/Fisher's linear discriminant analysisL1-normPrincipal component analysisinfo:eu-repo/semantics/articleinfo:eu-repo/semantics/openAccess10.1109/TSP.2019.2955860