Certifying and removing disparate impact
read the original abstract
What does it mean for an algorithm to be biased? In U.S. law, unintentional bias is encoded via disparate impact, which occurs when a selection process has widely different outcomes for different groups, even as it appears to be neutral. This legal determination hinges on a definition of a protected class (ethnicity, gender, religious practice) and an explicit description of the process. When the process is implemented using computers, determining disparate impact (and hence bias) is harder. It might not be possible to disclose the process. In addition, even if the process is open, it might be hard to elucidate in a legal setting how the algorithm makes its decisions. Instead of requiring access to the algorithm, we propose making inferences based on the data the algorithm uses. We make four contributions to this problem. First, we link the legal notion of disparate impact to a measure of classification accuracy that while known, has received relatively little attention. Second, we propose a test for disparate impact based on analyzing the information leakage of the protected class from the other data attributes. Third, we describe methods by which data might be made unbiased. Finally, we present empirical evidence supporting the effectiveness of our test for disparate impact and our approach for both masking bias and preserving relevant information in the data. Interestingly, our approach resembles some actual selection practices that have recently received legal scrutiny.
This paper has not been read by Pith yet.
Forward citations
Cited by 6 Pith papers
-
Toward Calibrated, Fair, and accurate Deepfake Detection
Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
-
Marginal Alignment Does Not Guarantee Joint-Distribution Fidelity: An Official-Reference Audit of Nemotron-Personas-Korea with Cross-Locale Replication
An audit of one million Korean synthetic personas shows marginal demographic alignment does not preserve joint distributions, with three specific mismatches identified via a new Independence-Assumption Footprint method.
-
Unmasking LAION-5B: Age, Gender, Race, and Emotion Biases in Large-Scale Image Datasets
Empirical audit of LAION-2B-en and LAION-2B-multi finds overrepresentation of young adults, White people, and males plus stereotypical emotion associations across two attribute classifiers.
-
A Blended Likelihood Approach for Achieving Fairness Using Naive Bayes
BMNB uses blended likelihood with tunable alpha and post-processing calibration to achieve DI close to 1 and improved EOD on three standard fairness datasets while preserving efficiency.
-
FAIR_XAI: Improving Multimodal Foundation Model Fairness via Explainability for Wellbeing Assessment
Vision-language models for wellbeing assessment exhibit dataset-dependent performance and demographic biases, with explainability interventions providing inconsistent fairness gains at potential accuracy costs.
-
Alternatives to the Laplacian for Scalable Spectral Clustering with Group Fairness Constraints
Fair-SMW uses SMW identity and alternative Laplacians to produce group-fair spectral clustering that is twice as fast and twice as balanced as prior methods on SBM and real network data.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.