Fast Bivariate Penalized Splines: the Sandwich Smoother

TL;DR

The paper introduces a fast, tensor product-based bivariate penalized spline smoothing method called the sandwich smoother, with proven asymptotic properties and superior computational efficiency for large datasets.

Contribution

It presents the first central limit theorem for a bivariate spline estimator and extends the method to higher-dimensional array data, significantly improving computational speed.

Findings

Sandwich smoother is orders of magnitude faster than existing methods.

It has comparable mean squared error to other smoothers.

The method is particularly effective for large functional data sets.

Abstract

We propose a fast penalized spline method for bivariate smoothing. Univariate P-spline smoothers (Eilers and Marx, 1996) are applied simultaneously along both coordinates. The new smoother has a sandwich form which suggested the name "sandwich smoother" to a referee. The sandwich smoother has a tensor product structure that simplifies an asymptotic analysis and it can be fast computed. We derive a local central limit theorem for the sandwich smoother, with simple expressions for the asymptotic bias and variance, by showing that the sandwich smoother is asymptotically equivalent to a bivariate kernel regression estimator with a product kernel. As far as we are aware, this is the first central limit theorem for a bivariate spline estimator of any type. Our simulation study shows that the sandwich smoother is orders of magnitude faster to compute than other bivariate spline smoothers, even when the latter are computed using a fast GLAM (Generalized Linear Array Model) algorithm, and comparable to them in terms of mean squared integrated errors. We extend the sandwich smoother to array data of higher dimensions, where a GLAM algorithm improves the computational speed of the sandwich smoother. One important application of the sandwich smoother is to estimate covariance functions in functional data analysis. In this application, our numerical results show that the sandwich smoother is orders of magnitude faster than local linear regression. The speed of the sandwich formula is important because functional data sets are becoming quite large.