Institutional Repository
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Reduced-rank L1-norm Principal-Component Analysis with performance guarantees

Kamrani Hossein, Asli Alireza Zolghadr, Markopoulos Panagiotis, Langberg Michael, Pados Dimitris A., Karystinos Georgios

Full record


URI: http://purl.tuc.gr/dl/dias/D364CA99-C7DD-4326-A237-65839F1851AB
Year 2021
Type of Item Peer-Reviewed Journal Publication
License
Details
Bibliographic Citation H. Kamrani, A. Z. Asli, P. P. Markopoulos, M. Langberg, D. A. Pados and G. N. Karystinos, "Reduced-rank L1-norm Principal-Component Analysis with performance guarantees," IEEE Trans. Signal Process., vol. 69, pp. 240-255, 2021, doi: 10.1109/TSP.2020.3039599. https://doi.org/10.1109/TSP.2020.3039599
Appears in Collections

Summary

Standard Principal-Component Analysis (PCA) is known to be sensitive to outliers among the processed data. On the other hand, L1-norm-based PCA (L1-PCA) exhibits sturdy resistance against outliers, while it performs similar to standard PCA when applied to nominal or smoothly corrupted data [1]. Exact calculation of the K L1-norm Principal Components (L1-PCs) of a rank-r datamatrix X ∈ℝ D×N costs O(N (r-1)K+1 ) [1], [2]. In this work, we present reduced-rank L1-PCA (RR L1-PCA): a hybrid approach that approximates the K L1-PCs of X by the L1-PCs of its L2-norm-based rank-d approximation (d ≤ r), calculable exactly with reduced complexity O(N (d-1)K+1 ). The proposed method combines the denoising capabilities and low computation cost of standard PCA with the outlier-resistance of L1-PCA. RR L1-PCA is accompanied by formal performance guarantees as well as thorough numerical studies that corroborate its computational and corruption resistance merits.

Services

Statistics