Eliminating Latent Discrimination: Train Then Mask

Ghili, Soheil; Kazemi, Ehsan; Karbasi, Amin

doi:10.2139/ssrn.3309776

Cited by 2 publications

(2 citation statements)

References 18 publications

(25 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The inclusion of sensitive data should be based on the potential for latent discrimination even in the absence of sensitive data, the relative availability and completeness of sensitive attributes, a priori knowledge of which sensitive features are responsible for bias, and many other related factors. 112 , 113 Uniformly defining which features should or should not be included in a model is overly restrictive. Our checklist was designed to give model developers a framework with which to discuss these sensitive yet important topics.…”

Section: Discussionmentioning

confidence: 99%

A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models

Wang

Landers

Adams³

et al. 2022

Journal of the American Medical Informatics Association

View full text Add to dashboard Cite

Objective Health care providers increasingly rely upon predictive algorithms when making important treatment decisions, however, evidence indicates that these tools can lead to inequitable outcomes across racial and socio-economic groups. In this study, we introduce a bias evaluation checklist that allows model developers and health care providers a means to systematically appraise a model’s potential to introduce bias. Materials and Methods Our methods include developing a bias evaluation checklist, a scoping literature review to identify 30-day hospital readmission prediction models, and assessing the selected models using the checklist. Results We selected 4 models for evaluation: LACE, HOSPITAL, Johns Hopkins ACG, and HATRIX. Our assessment identified critical ways in which these algorithms can perpetuate health care inequalities. We found that LACE and HOSPITAL have the greatest potential for introducing bias, Johns Hopkins ACG has the most areas of uncertainty, and HATRIX has the fewest causes for concern. Discussion Our approach gives model developers and health care providers a practical and systematic method for evaluating bias in predictive models. Traditional bias identification methods do not elucidate sources of bias and are thus insufficient for mitigation efforts. With our checklist, bias can be addressed and eliminated before a model is fully developed or deployed. Conclusion The potential for algorithms to perpetuate biased outcomes is not isolated to readmission prediction models; rather, we believe our results have implications for predictive models across health care. We offer a systematic method for evaluating potential bias with sufficient flexibility to be utilized across models and applications.

show abstract

Section: Discussionmentioning

confidence: 99%

A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models

Wang

Landers

Adams³

et al. 2022

Journal of the American Medical Informatics Association

View full text Add to dashboard Cite

show abstract

“…The work on this realm is focused on biased data or biased algorithms, however, using these biased algorithms in decision-making systems would lead into generating more biased data. This makes the causality of the fairness problem more complicated that exacerbates the problem even further (Barocas et al, 2017;Ghili et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

Efficient Fair Principal Component Analysis

Kamani¹,

Haddadpour²,

Forsati³

et al. 2019

Preprint

View full text Add to dashboard Cite

The flourishing assessments of fairness measure in machine learning algorithms have shown that dimension reduction methods such as PCA treat data from different sensitive groups unfairly. In particular, by aggregating data of different groups, the reconstruction error of the learned subspace becomes biased towards some populations that might hurt or benefit those groups inherently, leading to an unfair representation. On the other hand, alleviating the bias to protect sensitive groups in learning the optimal projection, would lead to a higher reconstruction error overall. This introduces a trade-off between sensitive groups' sacrifices and benefits, and the overall reconstruction error. In this paper, in pursuit of achieving fairness criteria in PCA, we introduce a more efficient notion of Pareto fairness, cast the Pareto fair dimensionality reduction as a multi-objective optimization problem, and propose an adaptive gradient-based algorithm to solve it. Using the notion of Pareto optimality, we can guarantee that the solution of our proposed algorithm belongs to the Pareto frontier for all groups, which achieves the optimal trade-off between those aforementioned conflicting objectives. This framework can be efficiently generalized to multiple group sensitive features, as well. We provide convergence analysis of our algorithm for both convex and non-convex objectives and show its efficacy through empirical studies on different datasets, in comparison with the state-of-the-art algorithm.

show abstract

Eliminating Latent Discrimination: Train Then Mask

Cited by 2 publications

References 18 publications

A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models

A bias evaluation checklist for predictive models and its pilot application for 30-day hospital readmission models

Efficient Fair Principal Component Analysis

Contact Info

Product

Resources

About