Enhancing Quantile Function Estimation with Beta-Kernel Smoothing
Preprint
- 20 February 2025
- preprint
- Published by Springer Nature in Research Square
Abstract
This paper introduces a class of nonparametric quantile function estimators based on Beta kernel smoothing. We conduct a rigorous investigation into their large-sample properties, including asymptotic normality and mean squared equivalence to existing methods. Through extensive simulation studies, we demonstrate that the proposed Beta kernel estimators perform comparably to or outperform traditional empirical and symmetric location-scale kernel-based quantile estimators. Additionally, we provide two real-world applications to illustrate the practical effectiveness of our approach. The results suggest that Beta kernel smoothing offers a flexible and efficient alternative for quantile function estimation, particularly in cases where classical methods exhibit inefficiencies.Keywords
All Related Versions
- Version 1, 2025-02-20, Research Square
- Published version: Statistics and Computing, 35 (6), 1.
This publication has 9 references indexed in Scilit:
- Bernstein-type bounds for beta distributionModern Stochastics: Theory and Applications, 2023
- Adaptive nonparametric regression on finite supportCommunications in Statistics - Theory and Methods, 2022
- Beta kernel quantile estimators of heavy-tailed loss distributionsStatistics and Computing, 2009
- Beta kernel estimators for density functionsComputational Statistics & Data Analysis, 1999
- Relative Deficiency of Kernel Type Estimators of QuantilesThe Annals of Statistics, 1984
- A normal approximation for beta and gamma tail probabilitiesProbability Theory and Related Fields, 1984
- Quantile interval estimationCommunications in Statistics - Theory and Methods, 1983
- Linear Functions of Order Statistics with Smooth Weight FunctionsThe Annals of Statistics, 1974
- The Mean-Square Error of Bahadur's Order-Statistic ApproximationThe Annals of Statistics, 1973