Attribute Removal (Fairness Through Unawareness)

Description

Attribute Removal (Fairness Through Unawareness) ensures fairness by completely excluding protected attributes such as race, gender, or age from the model's input features. While this approach prevents direct discrimination, it may not eliminate bias if other features are correlated with protected attributes (proxy discrimination). This technique represents the most basic fairness intervention but often needs to be combined with other approaches to address indirect bias through seemingly neutral features.

Example Use Cases

Fairness

Removing gender, race, and age attributes from hiring algorithms to prevent direct discrimination, whilst acknowledging that indirect bias may persist through correlated features like education institution or postal code.

Excluding protected demographic attributes from credit scoring models to comply with fair lending regulations, ensuring no explicit consideration of race, gender, or ethnicity in loan approval decisions.

Building medical diagnosis models that exclude patient race and ethnicity to prevent biased treatment recommendations, whilst ensuring clinical decisions are based solely on medical indicators and symptoms.

Transparency

Creating transparent regulatory reporting systems that demonstrate compliance by explicitly documenting which protected attributes have been excluded from decision-making algorithms, providing clear audit trails for regulatory review.

Limitations

Proxy discrimination remains a major concern as seemingly neutral features (education, postal code, previous employment) may strongly correlate with protected attributes, perpetuating indirect bias.
Intersectional bias cannot be addressed through simple attribute removal, as complex interactions between multiple demographic characteristics may create compounding discrimination effects.
Legal and regulatory compliance may be insufficient, as many jurisdictions require demonstrating disparate impact absence rather than simply removing protected attributes from models.
Identifying all potential proxy variables is practically impossible, especially with high-dimensional data where subtle correlations with protected attributes may exist in unexpected features.
Performance degradation may occur if removed attributes contain legitimate predictive information, creating tension between fairness objectives and model accuracy requirements.

Resources

Research Papers

Fairness Through Awareness

Cynthia Dwork et al.•Apr 20, 2011

We study fairness in classification, where individuals are classified, e.g., admitted to a university, and the goal is to prevent discrimination against individuals based on their membership in some group, while maintaining utility for the classifier (the university). The main conceptual contribution of this paper is a framework for fair classification comprising (1) a (hypothetical) task-specific metric for determining the degree to which individuals are similar with respect to the classification task at hand; (2) an algorithm for maximizing utility subject to the fairness constraint, that similar individuals are treated similarly. We also present an adaptation of our approach to achieve the complementary goal of "fair affirmative action," which guarantees statistical parity (i.e., the demographics of the set of individuals receiving any classification are the same as the demographics of the underlying population), while treating similar individuals as similarly as possible. Finally, we discuss the relationship of fairness to privacy: when fairness implies privacy, and how tools developed in the context of differential privacy may be applied to fairness.

Fairness Constraints: Mechanisms for Fair Classification

Muhammad Bilal Zafar et al.•Jul 19, 2015

Algorithmic decision making systems are ubiquitous across a wide variety of online as well as offline services. These systems rely on complex learning methods and vast amounts of data to optimize the service functionality, satisfaction of the end user and profitability. However, there is a growing concern that these automated decisions can lead, even in the absence of intent, to a lack of fairness, i.e., their outcomes can disproportionately hurt (or, benefit) particular groups of people sharing one or more sensitive attributes (e.g., race, sex). In this paper, we introduce a flexible mechanism to design fair classifiers by leveraging a novel intuitive measure of decision boundary (un)fairness. We instantiate this mechanism with two well-known classifiers, logistic regression and support vector machines, and show on real-world data that our mechanism allows for a fine-grained control on the degree of fairness, often at a small cost in terms of accuracy.

Software Packages

fairlearn

May 15, 2018

A Python package to assess and improve fairness of machine learning models.

Documentations

The Ethical Algorithm: The Science of Socially Aware Algorithm Design

Kearns, Michael and Roth, Aaron•Nov 1, 2019

Related Techniques

Name	Description	Assurance Goals
Sensitivity Analysis for Fairness	Sensitivity Analysis for Fairness systematically evaluates how model predictions change when sensitive attributes or their proxies are perturbed whilst holding other factors constant. The technique involves creating counterfactual instances by modifying potentially discriminatory features (race, gender, age) or their correlates (zip code, names, education institutions) and measuring the resulting prediction differences. This controlled perturbation approach quantifies the degree to which protected characteristics influence model decisions, helping detect both direct discrimination and indirect bias through proxy variables even when sensitive attributes are not explicitly used as model inputs.	Fairness
Fairness GAN	A data generation technique that employs Generative Adversarial Networks (GANs) to create fair synthetic datasets by learning to generate data representations that preserve utility whilst obfuscating protected attributes. Unlike traditional GANs, Fairness GANs incorporate fairness constraints into the training objective, ensuring that the generated data maintains statistical parity across demographic groups. The technique can be used for data augmentation to balance underrepresented groups or to create privacy-preserving synthetic datasets that remove demographic bias from training data.	Fairness Privacy Reliability
Bayesian Fairness Regularization	Bayesian Fairness Regularization incorporates fairness constraints into machine learning models through Bayesian methods, treating fairness as a prior distribution or regularization term. This approach includes techniques like Fair Bayesian Optimization that use constrained optimization to tune model hyperparameters whilst enforcing fairness constraints, and methods that add regularization terms to objective functions to penalize discriminatory predictions. The technique allows for probabilistic interpretation of fairness constraints and can account for uncertainty in both model parameters and fairness requirements.	Fairness Reliability
Out-of-Domain Detection	Out-of-domain (OOD) detection identifies user inputs that fall outside an AI system's intended domain or capabilities, enabling graceful handling rather than generating unreliable responses. This technique trains classifiers or uses embedding-based methods to recognise when queries require knowledge or capabilities the system doesn't possess. OOD detection enables systems to explicitly decline, redirect, or request clarification rather than hallucinating answers or applying inappropriate reasoning to unfamiliar domains.	Reliability Safety Transparency