Generating Private and Fair Long-Sequenced Longitudinal Healthcare Records


Generating long-sequenced longitudinal healthcare records is critical as it has numerous potential applications. Long-sequenced longitudinal data allow us to better understand and find patterns from the data. However, privacy concerns make it challenging to share the data, and real-world data is not bias-free. Generative Adversarial Networks (GAN) have been used to synthesize healthcare records, but the high dimensionality of these data makes them challenging to generate. From these motivations, we are working on a diffusion-based model that is capable of generating long-sequenced, fair, and private healthcare records.

The 35th Swedish Artificial Intelligence Society (SAIS'23) Annual Workshop