Skip to yearly menu bar Skip to main content


Poster

Good Semi-supervised Learning That Requires a Bad GAN

Zihang Dai · Zhilin Yang · Fan Yang · William Cohen · Ruslan Salakhutdinov

Pacific Ballroom #111

Keywords: [ Semi-Supervised Learning ] [ Adversarial Networks ]


Abstract:

Semi-supervised learning methods based on generative adversarial networks (GANs) obtained strong empirical results, but it is not clear 1) how the discriminator benefits from joint training with a generator, and 2) why good semi-supervised classification performance and a good generator cannot be obtained at the same time. Theoretically we show that given the discriminator objective, good semi-supervised learning indeed requires a bad generator, and propose the definition of a preferred generator. Empirically, we derive a novel formulation based on our analysis that substantially improves over feature matching GANs, obtaining state-of-the-art results on multiple benchmark datasets.

Live content is unavailable. Log in and register to view live content