Tools

LEADER 03105nam a2200421 i 4500

001 on1570278125;

003 OCoLC;

005 20260127041227.0;

006 m f d ;

007 cr unu||||||||;

008 260127s2025 ncu|||||obm 000 0|eng d;

040 a| EREb| enge| rdac| ERE;

035 a| (OCoLC)1570278125;

049 a| EREE;

100 1 a| Akinola, Kehinde Oluwasayoe| author?| UNAUTHORIZED;

245 10 a| Fairness Adequacy Test For Machine Learning Systems;

264 1 a| [Greenville, N.C.] :b| [East Carolina Universityc| 2025.];

300 a| 334 pages;

336 a| textb| txt2| rdacontent;

337 a| computerb| c2| rdamedia;

338 a| online resourceb| cr2| rdacarrier;

347 a| text fileb| application/pdfc| 6.55 MB2| rda;

538 a| System requirements: Adobe Reader.;

538 a| Mode of access: World Wide Web.;

502 b| S.M.c| East Carolina Universityd| 2025.;

504 a| Includes bibliographical references.;

520 3 a| As Machine Learning (ML) systems assume a larger role in decisions that affectpeople's lives, such as who receives a loan, access to healthcare, or early release fromprison, ensuring these systems are fair is more important than ever. However, currentfairness checks often miss the subtle and complex ways in which bias can appear inalgorithms.This dissertation tackles that gap by proposing a practical framework for testinghow well machine learning models meet fairness standards, especially across protectedgroups. Instead of relying solely on standard performance metrics, the approach com-bines statistical tools with stress testing techniques to uncover hidden or overlookedbiases.There are four main contributions. First, we introduce a fairness adequacy testusing metrics like Equal Opportunity Difference (EOD) and Equalised Odds Metrics(EOM) to examine disparities in error rates across groups. Second, we apply mutationtesting by altering sensitive features such as race or gender to see how model outputschange, helping assess fairness under different conditions. Third, we use permutationmethods to simulate edge cases and test how models respond to unusual or extremeinputs. Finally, we validate this approach with real-world case studies in areas likehealthcare, finance, and criminal justice, where fairness is especially critical.By offering a clear and testable way to evaluate fairness, this work aims to supportthe development of more trustworthy, accountable, and equitable AI systems;

653 a| Machine Learning , Fair* Adequacy, AI, Systems;

700 1 a| Srinivasan, Madhusudane| degree supervisor.?| UNAUTHORIZED;

710 2 a| East Carolina University=| ^A637090;

856 40 z| Access via ScholarShipu| http://hdl.handle.net/10342/14425;

949 a| Click on web addressw| ASISh| JOYNER101;

949 a| Click on web addressw| ASISh| HSL111;

994 a| C0b| ERE;

596 a| 1 4;

998 a| 6699112;

999 a| CLICK ON WEB ADDRESSw| ASISc| 1i| 6699112-1001l| JNETm| JOYNERr| Ys| Yt| JNE3ETDu| 1/27/2026x| ETDz| JERESOURCE;

999 a| CLICK ON WEB ADDRESSw| ASISc| 1i| 6699112-2001l| HSLELECm| HSLr| Ys| Yt| HEETDu| 1/27/2026x| ETDz| HERESOURCE;

Fairness Adequacy Test For Machine Learning Systems

Author/creator	Akinola, Kehinde Oluwasayo author
Other author	Srinivasan, Madhusudan degree supervisor.
Other author	East Carolina University
Format	Theses and dissertations
Publication	[Greenville, N.C.] : [East Carolina University 2025.]
Description	334 pages
Supplemental Content	Access via ScholarShip

Summary	As Machine Learning (ML) systems assume a larger role in decisions that affectpeople's lives, such as who receives a loan, access to healthcare, or early release fromprison, ensuring these systems are fair is more important than ever. However, currentfairness checks often miss the subtle and complex ways in which bias can appear inalgorithms.This dissertation tackles that gap by proposing a practical framework for testinghow well machine learning models meet fairness standards, especially across protectedgroups. Instead of relying solely on standard performance metrics, the approach com-bines statistical tools with stress testing techniques to uncover hidden or overlookedbiases.There are four main contributions. First, we introduce a fairness adequacy testusing metrics like Equal Opportunity Difference (EOD) and Equalised Odds Metrics(EOM) to examine disparities in error rates across groups. Second, we apply mutationtesting by altering sensitive features such as race or gender to see how model outputschange, helping assess fairness under different conditions. Third, we use permutationmethods to simulate edge cases and test how models respond to unusual or extremeinputs. Finally, we validate this approach with real-world case studies in areas likehealthcare, finance, and criminal justice, where fairness is especially critical.By offering a clear and testable way to evaluate fairness, this work aims to supportthe development of more trustworthy, accountable, and equitable AI systems
Dissertation note	S.M. East Carolina University 2025.
Bibliography note	Includes bibliographical references.
Technical details	System requirements: Adobe Reader.
Technical details	Mode of access: World Wide Web.

Availability

Library	Location	Call Number	Status	Item Actions
	Electronic Resources	Access Content Online	✔ Available

Fairness Adequacy Test For Machine Learning Systems

Click here for more information about this title

Availability