A method for using generative adversarial neural networks to make a synthetic dataset with differential privacy to enable data sharing. This implementation uses the SPRINT randomized trial of blood pressure lowering as the data source.
Invited editorial comment on a new approach to evaluating whether patients are similar to those actually enrolled in a clinical trial, as opposed to the inclusion and exclusion criteria.
A method for using generative adversarial neural networks to make a synthetic dataset with differential privacy to enable data sharing. This implementation uses the SPRINT randomized trial of blood pressure lowering as the data source.