Large deviations of divergence measures on partitions

Jan Beirlant, Luc Devroye, L. Györfi, Igor Vajda

Research output: Contribution to journalArticle

28 Citations (Scopus)

Abstract

We discuss Chernoff-type large deviation results for the total variation, the I-divergence errors, and the χ2-divergence errors on partitions. In contrast to the total variation and the I-divergence, the χ2-divergence has an unconventional large deviation rate. Applications to Bahadur efficiencies of goodness-of-fit tests based on these divergence measures for multivariate observations are given.

Original languageEnglish
Pages (from-to)1-16
Number of pages16
JournalJournal of Statistical Planning and Inference
Volume93
Issue number1-2
DOIs
Publication statusPublished - Feb 2001

Fingerprint

Divergence Measure
Large Deviations
Divergence
Partition
Total Variation
Bahadur Efficiency
Goodness of Fit Test
Large deviations

Keywords

  • χ -divergence
  • 62G10
  • Bahadur exact slope
  • Goodness-of-fit
  • I-divergence
  • Large deviations
  • Partitions
  • Total variation

ASJC Scopus subject areas

  • Statistics, Probability and Uncertainty
  • Applied Mathematics
  • Statistics and Probability

Cite this

Large deviations of divergence measures on partitions. / Beirlant, Jan; Devroye, Luc; Györfi, L.; Vajda, Igor.

In: Journal of Statistical Planning and Inference, Vol. 93, No. 1-2, 02.2001, p. 1-16.

Research output: Contribution to journalArticle

Beirlant, Jan ; Devroye, Luc ; Györfi, L. ; Vajda, Igor. / Large deviations of divergence measures on partitions. In: Journal of Statistical Planning and Inference. 2001 ; Vol. 93, No. 1-2. pp. 1-16.
@article{78226dc0118e4ec381231393a0e7fefe,
title = "Large deviations of divergence measures on partitions",
abstract = "We discuss Chernoff-type large deviation results for the total variation, the I-divergence errors, and the χ2-divergence errors on partitions. In contrast to the total variation and the I-divergence, the χ2-divergence has an unconventional large deviation rate. Applications to Bahadur efficiencies of goodness-of-fit tests based on these divergence measures for multivariate observations are given.",
keywords = "χ -divergence, 62G10, Bahadur exact slope, Goodness-of-fit, I-divergence, Large deviations, Partitions, Total variation",
author = "Jan Beirlant and Luc Devroye and L. Gy{\"o}rfi and Igor Vajda",
year = "2001",
month = "2",
doi = "10.1016/S0378-3758(00)00202-0",
language = "English",
volume = "93",
pages = "1--16",
journal = "Journal of Statistical Planning and Inference",
issn = "0378-3758",
publisher = "Elsevier",
number = "1-2",

}

TY - JOUR

T1 - Large deviations of divergence measures on partitions

AU - Beirlant, Jan

AU - Devroye, Luc

AU - Györfi, L.

AU - Vajda, Igor

PY - 2001/2

Y1 - 2001/2

N2 - We discuss Chernoff-type large deviation results for the total variation, the I-divergence errors, and the χ2-divergence errors on partitions. In contrast to the total variation and the I-divergence, the χ2-divergence has an unconventional large deviation rate. Applications to Bahadur efficiencies of goodness-of-fit tests based on these divergence measures for multivariate observations are given.

AB - We discuss Chernoff-type large deviation results for the total variation, the I-divergence errors, and the χ2-divergence errors on partitions. In contrast to the total variation and the I-divergence, the χ2-divergence has an unconventional large deviation rate. Applications to Bahadur efficiencies of goodness-of-fit tests based on these divergence measures for multivariate observations are given.

KW - χ -divergence

KW - 62G10

KW - Bahadur exact slope

KW - Goodness-of-fit

KW - I-divergence

KW - Large deviations

KW - Partitions

KW - Total variation

UR - http://www.scopus.com/inward/record.url?scp=0008239705&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0008239705&partnerID=8YFLogxK

U2 - 10.1016/S0378-3758(00)00202-0

DO - 10.1016/S0378-3758(00)00202-0

M3 - Article

AN - SCOPUS:0008239705

VL - 93

SP - 1

EP - 16

JO - Journal of Statistical Planning and Inference

JF - Journal of Statistical Planning and Inference

SN - 0378-3758

IS - 1-2

ER -