Methodology and convergence rates for functional linear regression

TL;DR

This paper analyzes methods for estimating the slope function in functional linear regression, comparing spectral decomposition with PCA and quadratic regularisation, and discusses their convergence rates and advantages.

Contribution

It provides a detailed analysis of PCA-based and quadratic regularisation methods, highlighting conditions for optimal convergence in functional linear regression.

Findings

PCA-based methods can achieve optimal convergence rates under certain conditions.

Quadratic regularisation offers advantages in specific scenarios.

The paper compares spectral decomposition and regularisation approaches in detail.

Abstract

In functional linear regression, the slope ``parameter'' is a function. Therefore, in a nonparametric context, it is determined by an infinite number of unknowns. Its estimation involves solving an ill-posed problem and has points of contact with a range of methodologies, including statistical smoothing and deconvolution. The standard approach to estimating the slope function is based explicitly on functional principal components analysis and, consequently, on spectral decomposition in terms of eigenvalues and eigenfunctions. We discuss this approach in detail and show that in certain circumstances, optimal convergence rates are achieved by the PCA technique. An alternative approach based on quadratic regularisation is suggested and shown to have advantages from some points of view.