Functional worst risk minimization

TL;DR

This paper extends worst risk minimization to the functional setting, allowing for unbounded operators and providing a new eigenfunction-free solution that simplifies estimation in functional regression under distribution shifts.

Contribution

It generalizes worst risk minimization to functional data, introduces conditions for minimizer existence without eigenfunction estimation, and offers consistent estimators with large sample bounds.

Findings

Decomposition of worst risk similar to non-functional case

Existence of minimizer characterized in square integrable kernels

Proposed estimators are consistent with large sample bounds

Abstract

The aim of this paper is to extend worst risk minimization, also called worst average loss minimization, to the functional realm. This means finding a functional regression representation that will be robust to future distribution shifts on the basis of data from two environments. In the classical non-functional realm, structural equations are based on a transfer matrix $B$. In section~\ref{sec:sfr}, we generalize this to consider a linear operator $\mathcal{T}$ on square integrable processes that plays the the part of $B$. By requiring that $(I-\mathcal{T})^{-1}$ is bounded -- as opposed to $\mathcal{T}$ -- this will allow for a large class of unbounded operators to be considered. Section~\ref{sec:worstrisk} considers two separate cases that both lead to the same worst-risk decomposition. Remarkably, this decomposition has the same structure as in the non-functional case. We consider any operator $\mathcal{T}$ that makes $(I-\mathcal{T})^{-1}$ bounded and define the future shift set in terms of the covariance functions of the shifts. In section~\ref{sec:minimizer}, we prove a necessary and sufficient condition for existence of a minimizer to this worst risk in the space of square integrable kernels. Previously, such minimizers were expressed in terms of the unknown eigenfunctions of the target and covariate integral operators (see for instance \cite{HeMullerWang} and \cite{YaoAOS}). This means that in order to estimate the minimizer, one must first estimate these unknown eigenfunctions. In contrast, the solution provided here will be expressed in any arbitrary ON-basis. This completely removes any necessity of estimating eigenfunctions. This pays dividends in section~\ref{sec:estimation}, where we provide a family of estimators, that are consistent with a large sample bound. Proofs of all the results are provided in the appendix.