Uniform error bound for PCA matrix denoising

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Tong, Xin T., Wang, Wanjie, Wang, Yuguan
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929476840980480
author Tong, Xin T.
Wang, Wanjie
Wang, Yuguan
author_facet Tong, Xin T.
Wang, Wanjie
Wang, Yuguan
contents Principal component analysis (PCA) is a simple and popular tool for processing high-dimensional data. We investigate its effectiveness for matrix denoising. We consider the clean data are generated from a low-dimensional subspace, but masked by independent high-dimensional sub-Gaussian noises with standard deviation $σ$. Under the low-rank assumption on the clean data with a mild spectral gap assumption, we prove that the distance between each pair of PCA-denoised data point and the clean data point is uniformly bounded by $O(σ\log n)$. To illustrate the spectral gap assumption, we show it can be satisfied when the clean data are independently generated with a non-degenerate covariance matrix. We then provide a general lower bound for the error of the denoised data matrix, which indicates PCA denoising gives a uniform error bound that is rate-optimal. Furthermore, we examine how the error bound impacts downstream applications such as clustering and manifold learning. Numerical results validate our theoretical findings and reveal the importance of the uniform error.
format Preprint
id arxiv_https___arxiv_org_abs_2306_12690
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Uniform error bound for PCA matrix denoising
Tong, Xin T.
Wang, Wanjie
Wang, Yuguan
Statistics Theory
Methodology
62H25(primary), 62H30, 62R30
Principal component analysis (PCA) is a simple and popular tool for processing high-dimensional data. We investigate its effectiveness for matrix denoising. We consider the clean data are generated from a low-dimensional subspace, but masked by independent high-dimensional sub-Gaussian noises with standard deviation $σ$. Under the low-rank assumption on the clean data with a mild spectral gap assumption, we prove that the distance between each pair of PCA-denoised data point and the clean data point is uniformly bounded by $O(σ\log n)$. To illustrate the spectral gap assumption, we show it can be satisfied when the clean data are independently generated with a non-degenerate covariance matrix. We then provide a general lower bound for the error of the denoised data matrix, which indicates PCA denoising gives a uniform error bound that is rate-optimal. Furthermore, we examine how the error bound impacts downstream applications such as clustering and manifold learning. Numerical results validate our theoretical findings and reveal the importance of the uniform error.
title Uniform error bound for PCA matrix denoising
topic Statistics Theory
Methodology
62H25(primary), 62H30, 62R30
url https://arxiv.org/abs/2306.12690