Plug-in bandwidth selection for kernel density estimation with discrete data

Chi Yang Chu, Daniel J. Henderson, Christopher Parmeter

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

This paper proposes plug-in bandwidth selection for kernel density estimation with discrete data via minimization of mean summed square error. Simulation results show that the plug-in bandwidths perform well, relative to cross-validated bandwidths, in non-uniform designs. We further find that plug-in bandwidths are relatively small. Several empirical examples show that the plug-in bandwidths are typically similar in magnitude to their cross-validated counterparts.

Original languageEnglish (US)
Pages (from-to)199-214
Number of pages16
JournalEconometrics
Volume3
Issue number2
DOIs
StatePublished - Jun 1 2015

Fingerprint

Kernel density estimation
Bandwidth
Simulation

Keywords

  • Bandwidth selection
  • Discrete variable
  • Kernel
  • Nonparametric
  • Plug-in

ASJC Scopus subject areas

  • Economics and Econometrics

Cite this

Plug-in bandwidth selection for kernel density estimation with discrete data. / Chu, Chi Yang; Henderson, Daniel J.; Parmeter, Christopher.

In: Econometrics, Vol. 3, No. 2, 01.06.2015, p. 199-214.

Research output: Contribution to journalArticle

Chu, Chi Yang ; Henderson, Daniel J. ; Parmeter, Christopher. / Plug-in bandwidth selection for kernel density estimation with discrete data. In: Econometrics. 2015 ; Vol. 3, No. 2. pp. 199-214.
@article{8c6f036bcf044fd6914d2946a22f3500,
title = "Plug-in bandwidth selection for kernel density estimation with discrete data",
abstract = "This paper proposes plug-in bandwidth selection for kernel density estimation with discrete data via minimization of mean summed square error. Simulation results show that the plug-in bandwidths perform well, relative to cross-validated bandwidths, in non-uniform designs. We further find that plug-in bandwidths are relatively small. Several empirical examples show that the plug-in bandwidths are typically similar in magnitude to their cross-validated counterparts.",
keywords = "Bandwidth selection, Discrete variable, Kernel, Nonparametric, Plug-in",
author = "Chu, {Chi Yang} and Henderson, {Daniel J.} and Christopher Parmeter",
year = "2015",
month = "6",
day = "1",
doi = "10.3390/econometrics3020199",
language = "English (US)",
volume = "3",
pages = "199--214",
journal = "Econometrics",
issn = "2225-1146",
publisher = "MDPI AG",
number = "2",

}

TY - JOUR

T1 - Plug-in bandwidth selection for kernel density estimation with discrete data

AU - Chu, Chi Yang

AU - Henderson, Daniel J.

AU - Parmeter, Christopher

PY - 2015/6/1

Y1 - 2015/6/1

N2 - This paper proposes plug-in bandwidth selection for kernel density estimation with discrete data via minimization of mean summed square error. Simulation results show that the plug-in bandwidths perform well, relative to cross-validated bandwidths, in non-uniform designs. We further find that plug-in bandwidths are relatively small. Several empirical examples show that the plug-in bandwidths are typically similar in magnitude to their cross-validated counterparts.

AB - This paper proposes plug-in bandwidth selection for kernel density estimation with discrete data via minimization of mean summed square error. Simulation results show that the plug-in bandwidths perform well, relative to cross-validated bandwidths, in non-uniform designs. We further find that plug-in bandwidths are relatively small. Several empirical examples show that the plug-in bandwidths are typically similar in magnitude to their cross-validated counterparts.

KW - Bandwidth selection

KW - Discrete variable

KW - Kernel

KW - Nonparametric

KW - Plug-in

UR - http://www.scopus.com/inward/record.url?scp=85028328469&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85028328469&partnerID=8YFLogxK

U2 - 10.3390/econometrics3020199

DO - 10.3390/econometrics3020199

M3 - Article

VL - 3

SP - 199

EP - 214

JO - Econometrics

JF - Econometrics

SN - 2225-1146

IS - 2

ER -