Bandwidth selection for nonparametric modal regression

TL;DR

This paper introduces two new bandwidth selection methods specifically designed for nonparametric modal regression, improving the accuracy of local mode estimation compared to traditional density estimation methods.

Contribution

The paper develops and evaluates two novel bandwidth selection techniques tailored for mode estimation in nonparametric regression, addressing limitations of existing methods.

Findings

Proposed methods outperform traditional bandwidth selectors in mode estimation accuracy.

Numerical studies demonstrate improved performance on synthetic and real data.

New methods are effective in practical nonparametric modal regression applications.

Abstract

In the context of estimating local modes of a conditional density based on kernel density estimators, we show that existing bandwidth selection methods developed for kernel density estimation are unsuitable for mode estimation. We propose two methods to select bandwidths tailored for mode estimation in the regression setting. Numerical studies using synthetic data and a real-life data set are carried out to demonstrate the performance of the proposed methods in comparison with several well received bandwidth selection methods for density estimation.