<ul data-eligibleForWebStory="false"><li>As the use of predictive machine learning algorithms increases, ensuring fairness across sensitive groups is crucial.</li><li>Proxy-sensitive attributes are proposed as a solution for enforcing fairness in the absence of complete sensitive group information.</li><li>This work explores the use of proxy-sensitive attributes for multiaccuracy and multicalibration, providing bounds on fairness violations and demonstrating mitigation strategies.</li><li>Experiments on real-world datasets show that approximate multiaccuracy and multicalibration can be achieved even when sensitive group data is missing or incomplete.</li></ul>

Multiaccuracy and Multicalibration via Proxy Groups

Discover more