Bt-GAN: Generating Fair Synthetic Healthdata via Bias-transforming Generative Adversarial Networks

arXiv:2404.13634v1 Announce Type: new
Abstract: Synthetic data generation offers a promising solution to enhance the usefulness of Electronic Healthcare Records (EHR) by generating realistic de-identified data. However, the existing literature primarily focuses on the quality of synthetic health data, neglecting the crucial aspect of fairness in downstream predictions. Consequently, models trained on synthetic EHR have faced criticism for producing biased outcomes in target tasks. These biases can arise from either spurious correlations between features or the failure of models to …

abstract adversarial arxiv bias cs.ai cs.lg data ehr electronic fair fairness gan generative generative adversarial networks health healthcare health data however literature networks predictions quality records solution synthetic synthetic data type via