Research Area:  Machine Learning

Abstract:

Generative models can be used for a wide range of tasks, and have the appealing ability to learn from both labelled and unlabelled data. In contrast, discriminative models cannot learn from unlabelled data, but tend to outperform their generative counterparts in supervised tasks. We develop a framework to jointly train deep generative and discriminative models, enjoying the benefits of both. The framework allows models to learn from labelled and unlabelled data, as well as naturally account for uncertainty in predictive distributions, providing the first Bayesian approach to semi-supervised learning with deep generative models. We demonstrate that our blended discriminative and generative models outperform purely generative models in both predictive performance and uncertainty calibration in a number of semi-supervised learning tasks.

Keywords:  
Probabilistic models
Semi-supervised learning
Variational autoencoders
Predictive uncertainty

Author(s) Name:  Jonathan Gordon, José Miguel Hernández-Lobato

Publisher name:  Elsevier

DOI:  10.1016/j.patcog.2019.107156

Volume Information:  Volume 100

Paper Link:   https://www.sciencedirect.com/science/article/pii/S003132031930456X