Fairness without the sensitive attribute via Causal Variational Autoencoder
This addresses fairness challenges in privacy-sensitive settings where sensitive attributes are not collected, offering a practical solution for applications under regulations like GDPR.
The paper tackles the problem of achieving fairness in machine learning models when sensitive attributes are unavailable, proposing a causal variational autoencoder (SRCVAE) to infer a proxy for sensitive information and using it for bias mitigation. The approach demonstrates significant improvements, achieving higher accuracy while maintaining the same level of fairness on two real datasets.
In recent years, most fairness strategies in machine learning models focus on mitigating unwanted biases by assuming that the sensitive information is observed. However this is not always possible in practice. Due to privacy purposes and var-ious regulations such as RGPD in EU, many personal sensitive attributes are frequently not collected. We notice a lack of approaches for mitigating bias in such difficult settings, in particular for achieving classical fairness objectives such as Demographic Parity and Equalized Odds. By leveraging recent developments for approximate inference, we propose an approach to fill this gap. Based on a causal graph, we rely on a new variational auto-encoding based framework named SRCVAE to infer a sensitive information proxy, that serve for bias mitigation in an adversarial fairness approach. We empirically demonstrate significant improvements over existing works in the field. We observe that the generated proxy's latent space recovers sensitive information and that our approach achieves a higher accuracy while obtaining the same level of fairness on two real datasets, as measured using com-mon fairness definitions.