A Margin Generative Adversarial Network (MarginGAN) is proposed for semi-supervised learning problems. Like Triple-GAN, the proposed MarginGAN consists of three components---a generator, a discriminator and a classifier, among which two forms of adversarial training arise. The discriminator is trained as usual to distinguish real examples from fake examples produced by the generator. The new feature is that the classifier attempts to increase the margin of real examples and to decrease the margin of fake examples. On the contrary, the purpose of the generator is yielding realistic and large-margin examples in order to fool the discriminator and the classifier simultaneously. Pseudo labels are used for generated and unlabeled examples in training. Our method is motivated by the success of large-margin classifiers and the recent viewpoint that good semi-supervised learning requires a ``bad'' GAN. Experiments on benchmark datasets testify that MarginGAN is orthogonal to several state-of-the-art methods, offering improved error rates and shorter training time as well.
Jinhao Dong (Xidian University)
Tong Lin (Peking University)
Tong Lin received the PhD degree in Applied Mathematics from Peking University in 2001. In 1999 and 2000, he was a visiting student at the Media Computing Group, Microsoft Research Asia. In 2002, he joined the Key Laboratory of Machine Perception at Peking University, China, where he is currently an associate professor. From 2004 to 2005, he was an exchange scholar at Seoul National University, Korea. From 2007 to 2008, he was an exchange scholar at UCSD Moores Cancer Center, CA, USA. His recent research interests are machine learning algorithms and theories.