An Infinite Dimensional Analysis of Kernel Principal Components

MaRDI QIDQ6320543zbMATH OpenFDO

Authors Palle Jorgensen, Sooran Kang, Myung-Sin Song, Feng Tian

Publication date 14 June 2019

Numerical approximation of eigenvalues and of other parts of the spectrum of ordinary differential operators (34L16) Computational methods for ergodic theory (approximation of invariant measures, computation of Lyapunov exponents, entropy, etc.) (37M25) General harmonic expansions, frames (42C15) Nontrigonometric harmonic analysis involving wavelets and other special systems (42C40) (Generalized) eigenfunction expansions of linear operators; rigged Hilbert spaces (47A70) Riesz operators; eigenvalue distributions; approximation numbers, (s)-numbers, Kolmogorov numbers, entropy numbers, etc. of operators (47B06) Linear operators in reproducing-kernel Hilbert spaces (including de Branges, de Branges-Rovnyak, and other structured spaces) (47B32) Factor analysis and principal components; correspondence analysis (62H25) Algorithms for approximation of functions (65D15) Numerical optimization and variational techniques (65K10) Numerical methods for wavelets (65T60)

Abstract: We study non-linear data-dimension reduction. We are motivated by the classical linear framework of Principal Component Analysis. In nonlinear case, we introduce instead a new kernel-Principal Component Analysis, manifold and feature space transforms. Our results extend earlier work for probabilistic Karhunen-Lo`eve transforms on compression of wavelet images. Our object is algorithms for optimization, selection of efficient bases, or components, which serve to minimize entropy and error; and hence to improve digital representation of images, and hence of optimal storage, and transmission. We prove several new theorems for data-dimension reduction. Moreover, with the use of frames in Hilbert space, and a new Hilbert-Schmidt analysis, we identify when a choice of Gaussian kernel is optimal.

This page was built for publication: An Infinite Dimensional Analysis of Kernel Principal Components

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6320543)