Publication detail

Maximum likelihood method for bandwidth selection in kernel conditional density estimate

POKOROVÁ, K., HOROVÁ, I.

Original Title

Maximum likelihood method for bandwidth selection in kernel conditional density estimate

Type

journal article in Web of Science

Language

English

Original Abstract

This paper discusses the kernel estimator of conditional density. A significant problem of kernel smoothing is bandwidth selection. The problem consists in the fact that optimal bandwidth depends on the unknown conditional and marginal density. This is the reason why some data-driven method needs to be applied. In this paper, we suggest a method for bandwidth selection based on a classical maximum likelihood approach. We consider a slight modification of the original method—the maximum likelihood method with one observation being left out. Applied to two types of conditional density estimators—to the Nadaraya–Watson and local linear estimator, the proposed method is compared with other known methods in a simulation study. Our aim is to compare the methods from different points of view, concentrating on the accuracy of the estimated bandwidths, on the final model quality measure, and on the computational time.

Keywords

kernel smoothing; conditional density; methods for bandwidth selection; leave-one-out maximum likelihood method

Authors

POKOROVÁ, K., HOROVÁ, I.

Released

2. 11. 2019

Publisher

Springer Verlag

Location

Berlin

ISBN

0943-4062

Periodical

COMPUTATIONAL STATISTICS & DATA ANALYSIS

Year of study

34

Number

4

State

Federal Republic of Germany

Pages from

1871

Pages to

1887

Pages count

16

URL

BibTex

@article{BUT159825,
  author="Kateřina {Pokorová} and Ivanka {Horová}",
  title="Maximum likelihood method for bandwidth selection in kernel conditional density estimate",
  journal="COMPUTATIONAL STATISTICS & DATA ANALYSIS",
  year="2019",
  volume="34",
  number="4",
  pages="1871--1887",
  doi="10.1007/s00180-019-00884-0",
  issn="0943-4062",
  url="https://link.springer.com/article/10.1007/s00180-019-00884-0"
}