Semi-Supervised Learning with Deep Generative Models

TL;DR

This paper demonstrates that deep generative models combined with variational inference significantly improve semi-supervised learning, enabling effective generalization from small labeled datasets to large unlabeled datasets.

Contribution

The authors develop scalable deep generative models with variational inference techniques that enhance semi-supervised learning performance over previous methods.

Findings

Deep generative models outperform traditional approaches in semi-supervised tasks.

Variational inference enables efficient training of complex generative models.

The proposed models are scalable and effective with limited labeled data.

Abstract

The ever-increasing size of modern data sets combined with the difficulty of obtaining label information has made semi-supervised learning one of the problems of significant practical importance in modern data analysis. We revisit the approach to semi-supervised learning with generative models and develop new models that allow for effective generalisation from small labelled data sets to large unlabelled ones. Generative approaches have thus far been either inflexible, inefficient or non-scalable. We show that deep generative models and approximate Bayesian inference exploiting recent advances in variational methods can be used to provide significant improvements, making generative approaches highly competitive for semi-supervised learning.