Disease Prediction from Electronic Health Records Using Generative Adversarial Networks

Electronic health records (EHRs) have contributed to the computerization of patient records so that they can be used not only for efficient and systematic medical services, but also for research on data science. In this paper, we compared the disease prediction performance of generative adversarial networks (GANs) and conventional learning algorithms in combination with missing value prediction methods. As a result, the highest accuracy of 98.05% was obtained using a stacked autoencoder as the missing value prediction method and an auxiliary classifier GANs (AC-GANs) as the disease predicting method. Our results show that the combination of the stacked autoencoder and the AC-GANs significantly outperforms existing algorithms for the problem of disease prediction in which missing values and class imbalance exist.