Sorry, you need to enable JavaScript to visit this website.

ROBUST PRINCIPAL COMPONENT ANALYSIS USING ALPHA DIVERGENCE

Citation Author(s):
Abd-Krim Seghouane
Submitted by:
Aref Miri Rekavandi
Last updated:
8 November 2020 - 7:05am
Document Type:
Presentation Slides
Document Year:
Event:
Presenters Name:
Aref Miri Rekavandi

Abstract 

Abstract: 

In this paper, a new robust principal component analysis (RPCA) method is proposed which enables us to exploit the main components of a given corrupted data with non-Gaussian outliers. The proposed method is based on the alpha-divergence which is a parametric measure from information geometry. The proposed method which is adjustable by the hyperparameter alpha, reduces to the classical PCA under certain parameters. In order to derive the main components, the alpha-divergence between the empirical data distribution and the assumed model is minimized with respect to the unknown parameters and then the singular value decomposition (SVD) of the estimated covariance matrix exploits the main direction of data. The proposed method is applied to some video and signal processing applications and the results show the superiority of the method over the classical PCA and other existing robust methods.

up
0 users have voted:

Dataset Files

ICIP2020RPCA.pdf

(39)