sdv-dev / TGAN

Generative adversarial training for generating synthetic tabular data.
MIT License
281 stars 91 forks source link

Usage of discriminator as classifier #61

Open enzodal opened 3 years ago

enzodal commented 3 years ago

I'm trying to "extract" the discriminator once the training of the GAN is over in order to use it as a classifier (ie. to give him new samples and to observe wether he recognizes it as a fake or a normal data). But I'm really struggling with the implementation. Is someone able to give me a shortcut in order to use the discriminator of the trained GAN ? Thanks