Unsupervised and Semi-supervised Learning with Categorical Generative Adversarial Networks

TL;DR

This paper introduces CatGAN, a novel framework combining generative adversarial networks and information maximization to learn robust classifiers from unlabeled or partially labeled data, with promising empirical results.

Contribution

It proposes a new objective function for semi-supervised learning that generalizes GANs and RIM, enhancing robustness and classification performance.

Findings

Demonstrates robustness on synthetic and image data

Qualitative analysis of generated samples

Links to discriminative clustering algorithms

Abstract

In this paper we present a method for learning a discriminative classifier from unlabeled or partially labeled data. Our approach is based on an objective function that trades-off mutual information between observed examples and their predicted categorical class distribution, against robustness of the classifier to an adversarial generative model. The resulting algorithm can either be interpreted as a natural generalization of the generative adversarial networks (GAN) framework or as an extension of the regularized information maximization (RIM) framework to robust classification against an optimal adversary. We empirically evaluate our method - which we dub categorical generative adversarial networks (or CatGAN) - on synthetic data as well as on challenging image classification tasks, demonstrating the robustness of the learned classifiers. We further qualitatively assess the fidelity of samples generated by the adversarial generator that is learned alongside the discriminative classifier, and identify links between the CatGAN objective and discriminative clustering algorithms (such as RIM).