On thresholds for robust goodness-of-fit tests

Jayakrishnan Unnikrishnan; Sean Meyn; Venugopal V. Veeravalli

doi:10.1109/CIG.2010.5592803

On thresholds for robust goodness-of-fit tests

Jayakrishnan Unnikrishnan, Sean Meyn, Venugopal V. Veeravalli

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Goodness-of-fit tests are statistical procedures used to test the hypothesis H₀ that a set of observations were drawn according to some given probability distribution. Decision thresholds used in goodness-of-fit tests are typically set for guaranteeing a target false-alarm probability. In many popular testing procedures results on the weak convergence of the test statistics are used for setting approximate thresholds when exact computation is infeasible. In this work, we study robust procedures for goodness-of-fit where accurate models are not available for the distribution of the observations under hypothesis H₀. We develop procedures for setting thresholds in two specific examples a robust version of the Kolmogorov-Smirnov test for continuous alphabets and a robust version of the Hoeffding test for finite alphabets.

Original language	English (US)
Title of host publication	2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings
DOIs	https://doi.org/10.1109/CIG.2010.5592803
State	Published - 2010
Event	2010 IEEE Information Theory Workshop, ITW 2010 - Dublin, Ireland Duration: Aug 30 2010 → Sep 3 2010

Publication series

Name	2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings

Other

Other	2010 IEEE Information Theory Workshop, ITW 2010
Country/Territory	Ireland
City	Dublin
Period	8/30/10 → 9/3/10

ASJC Scopus subject areas

Information Systems
Applied Mathematics

Online availability

10.1109/CIG.2010.5592803

Library availability

Discover UIUC Full Text

Cite this

@inproceedings{45fed6801ba845798fab7beeff01242b,

title = "On thresholds for robust goodness-of-fit tests",

abstract = "Goodness-of-fit tests are statistical procedures used to test the hypothesis H0 that a set of observations were drawn according to some given probability distribution. Decision thresholds used in goodness-of-fit tests are typically set for guaranteeing a target false-alarm probability. In many popular testing procedures results on the weak convergence of the test statistics are used for setting approximate thresholds when exact computation is infeasible. In this work, we study robust procedures for goodness-of-fit where accurate models are not available for the distribution of the observations under hypothesis H0. We develop procedures for setting thresholds in two specific examples a robust version of the Kolmogorov-Smirnov test for continuous alphabets and a robust version of the Hoeffding test for finite alphabets.",

author = "Jayakrishnan Unnikrishnan and Sean Meyn and Veeravalli, {Venugopal V.}",

year = "2010",

doi = "10.1109/CIG.2010.5592803",

language = "English (US)",

isbn = "9781424482641",

series = "2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings",

booktitle = "2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings",

note = "2010 IEEE Information Theory Workshop, ITW 2010 ; Conference date: 30-08-2010 Through 03-09-2010",

}

TY - GEN

T1 - On thresholds for robust goodness-of-fit tests

AU - Unnikrishnan, Jayakrishnan

AU - Meyn, Sean

AU - Veeravalli, Venugopal V.

PY - 2010

Y1 - 2010

N2 - Goodness-of-fit tests are statistical procedures used to test the hypothesis H0 that a set of observations were drawn according to some given probability distribution. Decision thresholds used in goodness-of-fit tests are typically set for guaranteeing a target false-alarm probability. In many popular testing procedures results on the weak convergence of the test statistics are used for setting approximate thresholds when exact computation is infeasible. In this work, we study robust procedures for goodness-of-fit where accurate models are not available for the distribution of the observations under hypothesis H0. We develop procedures for setting thresholds in two specific examples a robust version of the Kolmogorov-Smirnov test for continuous alphabets and a robust version of the Hoeffding test for finite alphabets.

AB - Goodness-of-fit tests are statistical procedures used to test the hypothesis H0 that a set of observations were drawn according to some given probability distribution. Decision thresholds used in goodness-of-fit tests are typically set for guaranteeing a target false-alarm probability. In many popular testing procedures results on the weak convergence of the test statistics are used for setting approximate thresholds when exact computation is infeasible. In this work, we study robust procedures for goodness-of-fit where accurate models are not available for the distribution of the observations under hypothesis H0. We develop procedures for setting thresholds in two specific examples a robust version of the Kolmogorov-Smirnov test for continuous alphabets and a robust version of the Hoeffding test for finite alphabets.

UR - http://www.scopus.com/inward/record.url?scp=80051940474&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80051940474&partnerID=8YFLogxK

U2 - 10.1109/CIG.2010.5592803

DO - 10.1109/CIG.2010.5592803

M3 - Conference contribution

AN - SCOPUS:80051940474

SN - 9781424482641

T3 - 2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings

BT - 2010 IEEE Information Theory Workshop, ITW 2010 - Proceedings

T2 - 2010 IEEE Information Theory Workshop, ITW 2010

Y2 - 30 August 2010 through 3 September 2010

ER -

On thresholds for robust goodness-of-fit tests

Abstract

Publication series

Other

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this