FA(IR)2MA-GLVQ – A hidden-feature-bias mitigation approach for fairness in classification learning based on generalized matrix learning vector quantization.
Saved in:
| Title: | FA(IR)2MA-GLVQ – A hidden-feature-bias mitigation approach for fairness in classification learning based on generalized matrix learning vector quantization. |
|---|---|
| Authors: | Kaden, Marika1 (AUTHOR) kaden1@hs-mittweida.de, Schubert, Ronny1 (AUTHOR), Voigt, Julius1 (AUTHOR), Reuss, Lynn1 (AUTHOR), Engelsberger, Alexander1 (AUTHOR), Lövdal, Sofie2,3 (AUTHOR), van den Brandhof, Elina L.2,4 (AUTHOR), Biehl, Michael2,5 (AUTHOR), Villmann, Thomas1,6 (AUTHOR) villmann@hs-mittweida.de |
| Source: | Neurocomputing. May2026, Vol. 678, pN.PAG-N.PAG. 1p. |
| Subjects: | Algorithmic bias, Classification, Machine learning, Decision making, Discrimination (Sociology), Data analysis, Vector quantization |
| Abstract: | Developing fair classification models is a crucial aspect of machine learning research. However, unintended distortion in training data - biased data - can lead to discriminatory decisions. In this paper, we developed a workflow for detecting and mitigating bias in data using a shallow, interpretable machine learning model: the Generalized Matrix Learning Vector Quantization. We extend the approach by a relevance-based analysis to identify and reduce bias in the data. Combining similarity metric adaptation and relevance-based analysis, we can develop fair classification models that minimize the influence of bias in the data. Our results demonstrate that this method is effective in reducing bias in classification models and therefore supports fair decision-making. • Model native derived solution for bias mitigation avoiding surrogate methods. • Workflow to examine a bias hypothesis and, if applicable, mitigate the bias in a transparent manner. • Mitigated bias model performance is evaluated for significant differences compared to the original models performance. [ABSTRACT FROM AUTHOR] |
| Copyright of Neurocomputing is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
| FullText | Text: Availability: 0 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 192259359 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: FA(IR)2MA-GLVQ – A hidden-feature-bias mitigation approach for fairness in classification learning based on generalized matrix learning vector quantization. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Kaden%2C+Marika%22">Kaden, Marika</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> kaden1@hs-mittweida.de</i><br /><searchLink fieldCode="AR" term="%22Schubert%2C+Ronny%22">Schubert, Ronny</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Voigt%2C+Julius%22">Voigt, Julius</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Reuss%2C+Lynn%22">Reuss, Lynn</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Engelsberger%2C+Alexander%22">Engelsberger, Alexander</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Lövdal%2C+Sofie%22">Lövdal, Sofie</searchLink><relatesTo>2,3</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22van+den+Brandhof%2C+Elina+L%2E%22">van den Brandhof, Elina L.</searchLink><relatesTo>2,4</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Biehl%2C+Michael%22">Biehl, Michael</searchLink><relatesTo>2,5</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Villmann%2C+Thomas%22">Villmann, Thomas</searchLink><relatesTo>1,6</relatesTo> (AUTHOR)<i> villmann@hs-mittweida.de</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22Neurocomputing%22">Neurocomputing</searchLink>. May2026, Vol. 678, pN.PAG-N.PAG. 1p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Algorithmic+bias%22">Algorithmic bias</searchLink><br /><searchLink fieldCode="DE" term="%22Classification%22">Classification</searchLink><br /><searchLink fieldCode="DE" term="%22Machine+learning%22">Machine learning</searchLink><br /><searchLink fieldCode="DE" term="%22Decision+making%22">Decision making</searchLink><br /><searchLink fieldCode="DE" term="%22Discrimination+%28Sociology%29%22">Discrimination (Sociology)</searchLink><br /><searchLink fieldCode="DE" term="%22Data+analysis%22">Data analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Vector+quantization%22">Vector quantization</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Developing fair classification models is a crucial aspect of machine learning research. However, unintended distortion in training data - biased data - can lead to discriminatory decisions. In this paper, we developed a workflow for detecting and mitigating bias in data using a shallow, interpretable machine learning model: the Generalized Matrix Learning Vector Quantization. We extend the approach by a relevance-based analysis to identify and reduce bias in the data. Combining similarity metric adaptation and relevance-based analysis, we can develop fair classification models that minimize the influence of bias in the data. Our results demonstrate that this method is effective in reducing bias in classification models and therefore supports fair decision-making. • Model native derived solution for bias mitigation avoiding surrogate methods. • Workflow to examine a bias hypothesis and, if applicable, mitigate the bias in a transparent manner. • Mitigated bias model performance is evaluated for significant differences compared to the original models performance. [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of Neurocomputing is the property of Elsevier B.V. and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.) |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=192259359 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1016/j.neucom.2026.133200 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 1 StartPage: N.PAG Subjects: – SubjectFull: Algorithmic bias Type: general – SubjectFull: Classification Type: general – SubjectFull: Machine learning Type: general – SubjectFull: Decision making Type: general – SubjectFull: Discrimination (Sociology) Type: general – SubjectFull: Data analysis Type: general – SubjectFull: Vector quantization Type: general Titles: – TitleFull: FA(IR)2MA-GLVQ – A hidden-feature-bias mitigation approach for fairness in classification learning based on generalized matrix learning vector quantization. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Kaden, Marika – PersonEntity: Name: NameFull: Schubert, Ronny – PersonEntity: Name: NameFull: Voigt, Julius – PersonEntity: Name: NameFull: Reuss, Lynn – PersonEntity: Name: NameFull: Engelsberger, Alexander – PersonEntity: Name: NameFull: Lövdal, Sofie – PersonEntity: Name: NameFull: van den Brandhof, Elina L. – PersonEntity: Name: NameFull: Biehl, Michael – PersonEntity: Name: NameFull: Villmann, Thomas IsPartOfRelationships: – BibEntity: Dates: – D: 14 M: 05 Text: May2026 Type: published Y: 2026 Identifiers: – Type: issn-print Value: 09252312 Numbering: – Type: volume Value: 678 Titles: – TitleFull: Neurocomputing Type: main |
| ResultId | 1 |
