Intrinsic Grassmann Averages for Online Linear, Robust and Nonlinear Subspace Learning

被引:7
作者
Chakraborty, Rudrasis [1 ,3 ]
Yang, Liu [4 ]
Hauberg, Soren [5 ]
Vemuri, Baba C. [2 ,3 ]
机构
[1] Univ Florida, Gainesville, FL 32611 USA
[2] Univ Florida, Dept Comp Informat Sci & Engn, Engn, Gainesville, FL 32611 USA
[3] Univ Florida, Dept Stat, Gainesville, FL 32611 USA
[4] Univ Calif Berkeley, Berkeley, CA 94720 USA
[5] Tech Univ Denmark, Sect Cognit Syst, DK-2800 Lyngby, Denmark
基金
欧洲研究理事会;
关键词
Online subspace learning; robust; PCA; kernel PCA; grassmann manifold; frechet mean; frechet median; PRINCIPAL COMPONENT ANALYSIS; CENTER-OF-MASS; SPARSE; PCA; ALGORITHMS; UNIQUENESS; TRACKING; IMAGE;
D O I
10.1109/TPAMI.2020.2992392
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Principal component analysis (PCA) and Kernel principal component analysis (KPCA) are fundamental methods in machine learning for dimensionality reduction. The former is a technique for finding this approximation in finite dimensions and the latter is often in an infinite dimensional reproducing Kernel Hilbert-space (RKHS). In this paper, we present a geometric framework for computing the principal linear subspaces in both (finite and infinite) situations as well as for the robust PCA case, that amounts to computing the intrinsic average on the space of all subspaces: the Grassmann manifold. Points on this manifold are defined as the subspaces spanned by K-tuples of observations. The intrinsic Grassmann average of these subspaces are shown to coincide with the principal components of the observations when they are drawn from a Gaussian distribution. We show similar results in the RKHS case and provide an efficient algorithm for computing the projection onto the this average subspace. The result is a method akin to KPCA which is substantially faster. Further, we present a novel online version of the KPCA using our geometric framework. Competitive performance of all our algorithms are demonstrated on a variety of real and synthetic data sets.
引用
收藏
页码:3904 / 3917
页数:14
相关论文
共 57 条
[1]   RIEMANNIAN Lp CENTER OF MASS: EXISTENCE, UNIQUENESS, AND CONVEXITY [J].
Afsari, Bijan .
PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY, 2011, 139 (02) :655-673
[2]   First Efficient Convergence for Streaming k-PCA: a Global, Gap-Free, and Near-Optimal Rate [J].
Allen-Zhu, Zeyuan ;
Li, Yuanzhi .
2017 IEEE 58TH ANNUAL SYMPOSIUM ON FOUNDATIONS OF COMPUTER SCIENCE (FOCS), 2017, :487-492
[3]  
[Anonymous], 1966, Applied mathematics series
[4]  
[Anonymous], 2017, FOURIER ANAL GROUPS
[5]  
Balzano L., 2010, 2010 48th Annual Allerton Conference on Communication, Control, and Computing (Allerton), P704, DOI 10.1109/ALLERTON.2010.5706976
[6]   Streaming PCA and Subspace Tracking: The Missing Data Case [J].
Balzano, Laura ;
Chi, Yuejie ;
Lu, Yue M. .
PROCEEDINGS OF THE IEEE, 2018, 106 (08) :1293-1310
[7]   Stochastic Gradient Descent on Riemannian Manifolds [J].
Bonnabel, Silvere .
IEEE TRANSACTIONS ON AUTOMATIC CONTROL, 2013, 58 (09) :2217-2229
[8]  
Boutsidis C., 2014, P 26 ANN ACM SIAM S, P887
[9]  
Boyd S., 2009, Convex Optimization, DOI DOI 10.1017/CBO9780511804441
[10]  
Brand M, 2002, LECT NOTES COMPUT SC, V2350, P707