Abstract:Functional linear and single-index models are core regression methods in functional data analysis and are widely used for performing regression in a wide range of applications when the covariates are random functions coupled with scalar responses. In the existing literature, however, the construction of associated estimators and the study of their theoretical properties is invariably carried out on a case-by-case basis for specific models under consideration. In this work, assuming the predictors are Gaussian processes, we provide a unified methodological and theoretical framework for estimating the index in functional linear, and its direction in single-index models. In the latter case, the proposed approach does not require the specification of the link function. In terms of methodology, we show that the reproducing kernel Hilbert space (RKHS) based functional linear least-squares estimator, when viewed through the lens of an infinite-dimensional Gaussian Stein's identity, also provides an estimator of the index of the single-index model. Theoretically, we characterize the convergence rates of the proposed estimators for both linear and single-index models. Our analysis has several key advantages: (i) it does not require restrictive commutativity assumptions for the covariance operator of the random covariates and the integral operator associated with the reproducing kernel; and (ii) the true index parameter can lie outside of the chosen RKHS, thereby allowing for index misspecification as well as for quantifying the degree of such index misspecification. Several existing results emerge as special cases of our analysis.

J-contractive operator valued functions, vector valued de Branges spaces and functional models

Vector valued de Branges spaces of entire functions based on pairs of Fredholm operator valued functions and functional model

Some aspects of vector valued de Branges spaces of entire functions

Operator-Valued Typically Real Functions Induced by a Contraction

Generalized de Branges-Rovnyak spaces

Highly Non-contractive Iterated Function Systems on Euclidean Space Can Have an Attractor

Nonextreme de Branges-Rovnyak spaces as models for contractions

Noncommutative reproducing kernel Hilbert spaces

Common Attractors of Generalized Hutchinson–Wardowski Contractive Operators

Completely hyperexpansive operators with finite rank defect operator and de Branges-Rovnyak spaces

Analytic Kramer sampling and quasi Lagrange-type interpolation in vector valued RKHS

Operator models for meromorphic functions of bounded type

RKH spaces of Brownian type defined by Cesàro-Hardy operators

A Category of Banach Space Functors

A CLASS OF MULTIVALENT ANALYTIC FUNCTIONS INVOLVING THE GENERALIZED JUNG-KIM-SRIVASTAVA OPERATOR

A field theoretic operator model and Cowen-Douglas class

Operator realizations of non-commutative analytic functions

Sufficient Conditions for Linear Operators Related to Confluent Hypergeometric Function and Generalized Bessel Function of the First Kind to Belong to a Certain Class of Analytic Functions

$J$-self-adjoint operators with $\mathcal{C}$-symmetries: extension theory approach

Functionals with extrema at reproducing kernels

Functional linear and single-index models: A unified approach via Gaussian Stein identity