te Marvelde, Pepijn (author) Generative Adversarial Networks (GANs) are a modern solution aiming to encourage public sharing of data, even if the data contains inherently private information, by generating synthetic data that looks like, but is not equal to, the data the GAN was trained on. However, GANs are prone to remembering samples from the training data, therefore...
bachelor thesis 2021