Optimal prediction for additive function-on-function regression

DOI10.1214/18-EJS1505MaRDI QIDQ1711592zbMATH OpenFDO

Authors Matthew Reimherr, Bahaeddine Taoufik, Bharath K. Sriperumbudur

Publication date 18 January 2019

Published in Electronic Journal of Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1708.03372, https://projecteuclid.org/euclid.ejs/1545382950

zbMATH Keywords

functional data analysisminimax convergencenonlinear regressionreproducing kernel Hilbert space

Mathematics Subject Classification ID

Hilbert spaces with reproducing kernels (= (proper) functional Hilbert spaces, including de Branges-Rovnyak and other structured spaces) (46E22) Nonparametric regression and quantile regression (62G08) Asymptotic properties of nonparametric inference (62G20) General nonlinear regression (62J02) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Applications of statistics to actuarial sciences and financial mathematics (62P05)

Abstract: As with classic statistics, functional regression models are invaluable in the analysis of functional data. While there are now extensive tools with accompanying theory available for linear models, there is still a great deal of work to be done concerning nonlinear models for functional data. In this work we consider the Additive Function-on-Function Regression model, a type of nonlinear model that uses an additive relationship between the functional outcome and functional covariate. We present an estimation methodology built upon Reproducing Kernel Hilbert Spaces, and establish optimal rates of convergence for our estimates in terms of prediction error. We also discuss computational challenges that arise with such complex models, developing a representer theorem for our estimate as well as a more practical and computationally efficient approximation. Simulations and an application to Cumulative Intraday Returns around the 2008 financial crisis are also provided.

Review text

The paper under review deals with the problem of developing a penalized regression framework, based on Reproducing Kernel Hilbert Spaces (RKHS), for fitting the additive function-on-function regression model. The authors suggest an estimation methodology and establish optimal rates of convergence for their estimates in terms of prediction error. They also discuss computational aspects of the presented approach, ``as the RKHS structure allows for a fairly efficient computation as compared to the trivariate tensor product bases that have been used previously. Finally, simulations and an application to financial data are given.

Recommendations

• Additive functional regression in reproducing kernel Hilbert spaces under smoothness condition
• Non-asymptotic error bound for optimal prediction of function-on-function regression by RKHS approach
• Optimal Penalized Function-on-Function Regression Under a Reproducing Kernel Hilbert Space Framework
• Optimal prediction for high-dimensional functional quantile regression in reproducing kernel Hilbert spaces
• Structured Functional Additive Regression in Reproducing Kernel Hilbert Spaces

Cites work

• Additive Function-on-Function Regression
• Continuously additive models for nonlinear functional regression
• Continuously dynamic additive models for functional data
• Determining the order of the functional autoregressive model
• Fast covariance estimation for sparse functional data
• From sparse to dense functional data and beyond
• Functional Adaptive Model Estimation
• Functional additive mixed models
• Functional additive regression
• Functional Data Analysis for Sparse Longitudinal Data
• Functional Data Analysis with R and MATLAB
• Functional data analysis.
• Hilbert space methods in signal processing
• scientific article; zbMATH DE number 47282 (Why is no real title available?)
• scientific article; zbMATH DE number 5055767 (Why is no real title available?)
• Inference for functional data with applications
• Minimax and adaptive prediction for functional linear regression
• Nonlinear functional models for functional responses in reproducing kernel hilbert spaces
• Optimal estimation of the mean function based on discretely sampled functional data: phase transition
• Optimal Penalized Function-on-Function Regression Under a Reproducing Kernel Hilbert Space Framework
• Optimal Prediction in an Additive Functional Model
• Penalized likelihood functional regression
• Predictability of shapes of intraday price curves
• Regression models for functional data by reproducing kernel Hilbert spaces methods
• Some results on Tchebycheffian spline functions and stochastic processes
• Structured Functional Additive Regression in Reproducing Kernel Hilbert Spaces
• Testing stationarity of functional time series
• Tests for error correlation in the functional linear model
• Uniform convergence rates for nonparametric regression and principal component analysis in functional/longitudinal data

Cited in

(14)

• Optimal prediction for high-dimensional functional quantile regression in reproducing kernel Hilbert spaces
• Additive functional regression in reproducing kernel Hilbert spaces under smoothness condition
• Additive regression for non-Euclidean responses and predictors
• Non-asymptotic error bound for optimal prediction of function-on-function regression by RKHS approach
• Functional sufficient dimension reduction through average Fréchet derivatives
• Functional response additive model estimation with online virtual stock markets
• Optimal Prediction in an Additive Functional Model
• Estimation and Model Selection for Nonparametric Function-on-Function Regression
• Nonlinear function-on-function additive model with multiple predictor curves
• Low-Rank Covariance Function Estimation for Multidimensional Functional Data
• Optimal Function-on-Function Regression With Interaction Between Functional Predictors
• Modern nonlinear function-on-function regression
• Nonlinear Functional Modeling Using Neural Networks
• General Nonlinear Function-on-Function Regression via Functional Universal Approximation

Describes a project that uses

Uses Software

• R
• fda (R)
• tensorA

This page was built for publication: Optimal prediction for additive function-on-function regression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1711592)