Details
| Original language | English |
|---|---|
| Pages (from-to) | 2737-2770 |
| Number of pages | 34 |
| Journal | Knowledge and information systems |
| Volume | 64 |
| Issue number | 10 |
| Early online date | 27 Jul 2022 |
| Publication status | Published - Oct 2022 |
Abstract
Data-driven AI systems can lead to discrimination on the basis of protected attributes like gender or race. One cause for this is the encoded societal biases in the training data (e.g., under-representation of females in the tech workforce), which is aggravated in the presence of unbalanced class distributions (e.g., when “hired” is the minority class in a hiring application). State-of-the-art fairness-aware machine learning approaches focus on preserving the overall classification accuracy while mitigating discrimination. In the presence of class-imbalance, such methods may further aggravate the problem of discrimination by denying an already underrepresented group (e.g., females) the fundamental rights of equal social privileges (e.g., equal access to employment). To this end, we propose AdaFair, a fairness-aware boosting ensemble that changes the data distribution at each round, taking into account not only the class errors but also the fairness-related performance of the model defined cumulatively based on the partial ensemble. Except for the in-training boosting of the group discriminated over each round, AdaFair directly tackles imbalance during the post-training phase by optimizing the number of ensemble learners for balanced error performance. AdaFair can facilitate different parity-based fairness notions and mitigate effectively discriminatory outcomes.
Keywords
- Boosting, Class-imbalance, Disparate mistreatment, Ensemble learning, Equal opportunity, Fairness-aware classification, Statistical parity
ASJC Scopus subject areas
- Computer Science(all)
- Software
- Computer Science(all)
- Information Systems
- Computer Science(all)
- Human-Computer Interaction
- Computer Science(all)
- Hardware and Architecture
- Computer Science(all)
- Artificial Intelligence
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Knowledge and information systems, Vol. 64, No. 10, 10.2022, p. 2737-2770.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Parity-based cumulative fairness-aware boosting
AU - Iosifidis, Vasileios
AU - Roy, Arjun
AU - Ntoutsi, Eirini
N1 - Funding Information: The work is supported by the Volkswagen Foundation project BIAS (“Bias and Discrimination in Big Data and Algorithmic Processing. Philosophical Assessments, Legal Dimensions, and Technical Solutions”) within the initiative “AI and the Society of the Future”.
PY - 2022/10
Y1 - 2022/10
N2 - Data-driven AI systems can lead to discrimination on the basis of protected attributes like gender or race. One cause for this is the encoded societal biases in the training data (e.g., under-representation of females in the tech workforce), which is aggravated in the presence of unbalanced class distributions (e.g., when “hired” is the minority class in a hiring application). State-of-the-art fairness-aware machine learning approaches focus on preserving the overall classification accuracy while mitigating discrimination. In the presence of class-imbalance, such methods may further aggravate the problem of discrimination by denying an already underrepresented group (e.g., females) the fundamental rights of equal social privileges (e.g., equal access to employment). To this end, we propose AdaFair, a fairness-aware boosting ensemble that changes the data distribution at each round, taking into account not only the class errors but also the fairness-related performance of the model defined cumulatively based on the partial ensemble. Except for the in-training boosting of the group discriminated over each round, AdaFair directly tackles imbalance during the post-training phase by optimizing the number of ensemble learners for balanced error performance. AdaFair can facilitate different parity-based fairness notions and mitigate effectively discriminatory outcomes.
AB - Data-driven AI systems can lead to discrimination on the basis of protected attributes like gender or race. One cause for this is the encoded societal biases in the training data (e.g., under-representation of females in the tech workforce), which is aggravated in the presence of unbalanced class distributions (e.g., when “hired” is the minority class in a hiring application). State-of-the-art fairness-aware machine learning approaches focus on preserving the overall classification accuracy while mitigating discrimination. In the presence of class-imbalance, such methods may further aggravate the problem of discrimination by denying an already underrepresented group (e.g., females) the fundamental rights of equal social privileges (e.g., equal access to employment). To this end, we propose AdaFair, a fairness-aware boosting ensemble that changes the data distribution at each round, taking into account not only the class errors but also the fairness-related performance of the model defined cumulatively based on the partial ensemble. Except for the in-training boosting of the group discriminated over each round, AdaFair directly tackles imbalance during the post-training phase by optimizing the number of ensemble learners for balanced error performance. AdaFair can facilitate different parity-based fairness notions and mitigate effectively discriminatory outcomes.
KW - Boosting
KW - Class-imbalance
KW - Disparate mistreatment
KW - Ensemble learning
KW - Equal opportunity
KW - Fairness-aware classification
KW - Statistical parity
UR - http://www.scopus.com/inward/record.url?scp=85137649535&partnerID=8YFLogxK
U2 - 10.48550/arXiv.2201.01148
DO - 10.48550/arXiv.2201.01148
M3 - Article
AN - SCOPUS:85137649535
VL - 64
SP - 2737
EP - 2770
JO - Knowledge and information systems
JF - Knowledge and information systems
SN - 0219-1377
IS - 10
ER -