Fast and stable multivariate kernel density estimation by fast sum updating

TL;DR

This paper extends the fast sum updating method for kernel density estimation to multivariate data, improving speed and stability for large-scale problems with various kernels and adaptive bandwidths.

Contribution

It introduces a multivariate extension of the fast sum updating approach, applicable to general data, kernels, and adaptive bandwidths, enhancing efficiency and accuracy.

Findings

Method is fast, accurate, and stable in multivariate regression and density estimation.

Extension to multiple kernels and adaptive bandwidths improves versatility.

Numerical tests confirm the method's effectiveness for large datasets.

Abstract

Kernel density estimation and kernel regression are powerful but computationally expensive techniques: a direct evaluation of kernel density estimates at $M$ evaluation points given $N$ input sample points requires a quadratic $\mathcal{O}(MN)$ operations, which is prohibitive for large scale problems. For this reason, approximate methods such as binning with Fast Fourier Transform or the Fast Gauss Transform have been proposed to speed up kernel density estimation. Among these fast methods, the Fast Sum Updating approach is an attractive alternative, as it is an exact method and its speed is independent of the input sample and the bandwidth. Unfortunately, this method, based on data sorting, has for the most part been limited to the univariate case. In this paper, we revisit the fast sum updating approach and extend it in several ways. Our main contribution is to extend it to the general multivariate case for general input data and rectilinear evaluation grid. Other contributions include its extension to a wider class of kernels, including the triangular, cosine and Silverman kernels, its combination with parsimonious additive multivariate kernels, and its combination with a fast approximate k-nearest-neighbors bandwidth for multivariate datasets. Our numerical tests of multivariate regression and density estimation confirm the speed, accuracy and stability of the method. We hope this paper will renew interest for the fast sum updating approach and help solve large-scale practical density estimation and regression problems.