Unsupervised Learning Using Generative Adversarial Training And Clustering

Vittal Premachandran, Alan L. Yuille

Nov 05, 2016 (modified: Nov 05, 2016) ICLR 2017 conference submission readers: everyone
  • Abstract: In this paper, we propose an unsupervised learning approach that makes use of two components; a deep hierarchical feature extractor, and a more traditional clustering algorithm. We train the feature extractor in a purely unsupervised manner using generative adversarial training and, in the process, study the strengths of learning using a generative model as an adversary. We also show that adversarial training as done in Generative Adversarial Networks (GANs) is not sufficient to automatically group data into categorical clusters. Instead, we use a more traditional grouping algorithm, k-means clustering, to cluster the features learned using adversarial training. We experiment on three well-known datasets, CIFAR-10, CIFAR-100 and STL-10. The experiments show that the proposed approach performs similarly to supervised learning approaches, and, might even be better in situations with small amounts of labeled training data and large amounts of unlabeled data.
  • Conflicts: jhu.edu