Robust Subspace Tracking With Contamination Mitigation via  α-Divergence

Thanh, Le Trung; Miri Rekavandi, Aref; SEGHOUANE, ABD-KRIM; abed-meraim, karim

doi:10.36227/techrxiv.21385335.v2

2023_aFAPI.pdf (806.35 kB)

Robust Subspace Tracking With Contamination Mitigation via α-Divergence

preprint

posted on 2023-02-21, 20:01 authored by Le Trung ThanhLe Trung Thanh, Aref Miri Rekavandi, ABD-KRIM SEGHOUANEABD-KRIM SEGHOUANE, karim abed-meraim

Abstract: We studied the problem of robust subspace tracking (RST) in contaminated environments. Leveraging the fast approximated power iteration and α-divergence, a novel robust algorithm called αFAPI was developed for tracking the underlying principal subspace of streaming data over time. αFAPI is fast and it outperforms many RST methods while only having a low complexity linear to the data dimension. Some experiments were conducted to illustrate the performance of αFAPI.

Matlab Code: https://github.com/thanhtbt/aFAPI

Comment: Accepted at IEEE ICASSP 2023