Revisiting the predictive power of kernel principal components

被引：0

作者：

Jones, Ben ^{[1
]}

Artemiou, Andreas ^{[1
]}

机构：

[1] Cardiff Univ, Sch Math, Cardiff, S Glam, Wales

来源：

STATISTICS & PROBABILITY LETTERS | 2021年 / 171卷

基金：

英国工程与自然科学研究理事会;

关键词：

Model-free regression; Conditional independence; Hilbert spaces; Dimension reduction; Principal components; DIMENSION REDUCTION; FISHER LECTURE; REGRESSION;

D O I：

10.1016/j.spl.2020.109019

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

In this short note, recent results on the predictive power of kernel principal component in a regression setting are extended in two ways: (1) in the model-free setting, we relax a conditional independence model assumption to obtain a stronger result; and (2) the model-free setting is also extended in the infinite-dimensional setting. (C) 2020 The Author(s). Published by Elsevier B.V.

引用

页数：5

共 11 条

[1]

Artemiou A., 2021, USING MUTUAL INFORM

[2] Predictive power of principal components for single-index model and sufficient dimension reduction [J].