A composite likelihood cross-validation approach in selecting bandwidth for the estimation of the pair correlation function

Research output: Contribution to journalArticle

16 Citations (Scopus)

Abstract

A useful tool while analysing spatial point patterns is the pair correlation function (e.g. Fractals, Random Shapes and Point Fields, Wiley, New York, 1994). In practice, this function is often estimated by some nonparametric procedure such as kernel smoothing, where the smoothing parameter (i.e. bandwidth) is often determined arbitrarily. In this article, a data-driven method for the selection of the bandwidth is proposed. The efficacy of the proposed approach is studied through both simulations and an application to a forest data example.

Original languageEnglish (US)
Pages (from-to)336-346
Number of pages11
JournalScandinavian Journal of Statistics
Volume34
Issue number2
DOIs
StatePublished - Jun 2007
Externally publishedYes

Fingerprint

Composite Likelihood
Pair Correlation Function
Cross-validation
Bandwidth
Spatial Point Pattern
Random Fractals
Kernel Smoothing
Smoothing Parameter
Data-driven
Efficacy
Simulation

Keywords

  • Composite likelihood
  • Pair correlation function
  • Spatial point process

ASJC Scopus subject areas

  • Mathematics(all)
  • Statistics and Probability

Cite this

@article{8219054587544c23bea5229ea74eb37f,
title = "A composite likelihood cross-validation approach in selecting bandwidth for the estimation of the pair correlation function",
abstract = "A useful tool while analysing spatial point patterns is the pair correlation function (e.g. Fractals, Random Shapes and Point Fields, Wiley, New York, 1994). In practice, this function is often estimated by some nonparametric procedure such as kernel smoothing, where the smoothing parameter (i.e. bandwidth) is often determined arbitrarily. In this article, a data-driven method for the selection of the bandwidth is proposed. The efficacy of the proposed approach is studied through both simulations and an application to a forest data example.",
keywords = "Composite likelihood, Pair correlation function, Spatial point process",
author = "Yongtao Guan",
year = "2007",
month = "6",
doi = "10.1111/j.1467-9469.2006.00533.x",
language = "English (US)",
volume = "34",
pages = "336--346",
journal = "Scandinavian Journal of Statistics",
issn = "0303-6898",
publisher = "Wiley-Blackwell",
number = "2",

}

TY - JOUR

T1 - A composite likelihood cross-validation approach in selecting bandwidth for the estimation of the pair correlation function

AU - Guan, Yongtao

PY - 2007/6

Y1 - 2007/6

N2 - A useful tool while analysing spatial point patterns is the pair correlation function (e.g. Fractals, Random Shapes and Point Fields, Wiley, New York, 1994). In practice, this function is often estimated by some nonparametric procedure such as kernel smoothing, where the smoothing parameter (i.e. bandwidth) is often determined arbitrarily. In this article, a data-driven method for the selection of the bandwidth is proposed. The efficacy of the proposed approach is studied through both simulations and an application to a forest data example.

AB - A useful tool while analysing spatial point patterns is the pair correlation function (e.g. Fractals, Random Shapes and Point Fields, Wiley, New York, 1994). In practice, this function is often estimated by some nonparametric procedure such as kernel smoothing, where the smoothing parameter (i.e. bandwidth) is often determined arbitrarily. In this article, a data-driven method for the selection of the bandwidth is proposed. The efficacy of the proposed approach is studied through both simulations and an application to a forest data example.

KW - Composite likelihood

KW - Pair correlation function

KW - Spatial point process

UR - http://www.scopus.com/inward/record.url?scp=34547382412&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34547382412&partnerID=8YFLogxK

U2 - 10.1111/j.1467-9469.2006.00533.x

DO - 10.1111/j.1467-9469.2006.00533.x

M3 - Article

AN - SCOPUS:34547382412

VL - 34

SP - 336

EP - 346

JO - Scandinavian Journal of Statistics

JF - Scandinavian Journal of Statistics

SN - 0303-6898

IS - 2

ER -