Maximum likelihood method for bandwidth selection in kernel
conditional density estimate

J 2019

Maximum likelihood method for bandwidth selection in kernel conditional density estimate

POKOROVÁ, Kateřina and Ivanka HOROVÁ

Basic information

Original name

Maximum likelihood method for bandwidth selection in kernel conditional density estimate

Authors

POKOROVÁ, Kateřina (203 Czech Republic, belonging to the institution) and Ivanka HOROVÁ (203 Czech Republic, belonging to the institution)

Edition

Computational Statistics, Heidelberg, Springer Heidelberg, 2019, 0943-4062

Other information

Language

English

Type of outcome

Článek v odborném periodiku

Field of Study

10103 Statistics and probability

Country of publisher

Germany

Confidentiality degree

není předmětem státního či obchodního tajemství

References:

Full Text

Impact factor

Impact factor: 0.744

RIV identification code

RIV/00216224:14310/19:00111007

Organization unit

Faculty of Science

DOI

http://dx.doi.org/10.1007/s00180-019-00884-0

UT WoS

000501848900019

Keywords (in Czech)

jádrové vyhlazování; podmíněná hustota; metody odhadu vyhlazovacích parametrů; leave-one-out metoda maximální věrohodnosti

Keywords in English

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

Abstract

V originále

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.

Links

MUNI/A/1503/2018, interní kód MU

Name: Matematické statistické modelování 3 (Acronym: MaStaMo3)

Investor: Masaryk University, Category A

Citovat

POKOROVÁ, Kateřina and Ivanka HOROVÁ. Maximum likelihood method for bandwidth selection in kernel conditional density estimate. Computational Statistics. Heidelberg: Springer Heidelberg, 2019, vol. 34, No 4, p. 1871-1887. ISSN 0943-4062. Available from: https://dx.doi.org/10.1007/s00180-019-00884-0.

@article{1571859,
   author = {Pokorová, Kateřina and Horová, Ivanka},
   article_location = {Heidelberg},
   article_number = {4},
   doi = {http://dx.doi.org/10.1007/s00180-019-00884-0},
   keywords = {kernel smoothing; conditional density; methods for bandwidth selection; leave-one-out maximum likelihood method},
   language = {eng},
   issn = {0943-4062},
   journal = {Computational Statistics},
   title = {Maximum likelihood method for bandwidth selection in kernel conditional density estimate},
   url = {https://link.springer.com/article/10.1007/s00180-019-00884-0},
   volume = {34},
   year = {2019}
}

TY  - JOUR
ID  - 1571859
AU  - Pokorová, Kateřina - Horová, Ivanka
PY  - 2019
TI  - Maximum likelihood method for bandwidth selection in kernel conditional density estimate
JF  - Computational Statistics
VL  - 34
IS  - 4
SP  - 1871-1887
EP  - 1871-1887
PB  - Springer Heidelberg
SN  - 09434062
KW  - kernel smoothing
KW  - conditional density
KW  - methods for bandwidth selection
KW  - leave-one-out maximum likelihood method
UR  - https://link.springer.com/article/10.1007/s00180-019-00884-0
L2  - https://link.springer.com/article/10.1007/s00180-019-00884-0
N2  - 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.
ER  -

POKOROVÁ, Kateřina and Ivanka HOROVÁ. Maximum likelihood method for bandwidth selection in kernel conditional density estimate. \textit{Computational Statistics}. Heidelberg: Springer Heidelberg, 2019, vol.~34, No~4, p.~1871-1887. ISSN~0943-4062. Available from: https://dx.doi.org/10.1007/s00180-019-00884-0.

Detailed Information on Publication Record