Semi-Conditional Normalizing Flows for Semi-Supervised Learning

TL;DR

This paper introduces a semi-conditional normalizing flow model that effectively leverages both labeled and unlabeled data for semi-supervised learning, demonstrating superior performance over variational auto-encoders on MNIST.

Contribution

The paper presents a novel semi-conditional normalizing flow architecture with a conditional coupling layer for improved semi-supervised classification.

Findings

Outperforms variational auto-encoders on MNIST

Efficient computation of marginal likelihood gradients

Effective use of unlabeled data in the model

Abstract

This paper proposes a semi-conditional normalizing flow model for semi-supervised learning. The model uses both labelled and unlabeled data to learn an explicit model of joint distribution over objects and labels. Semi-conditional architecture of the model allows us to efficiently compute a value and gradients of the marginal likelihood for unlabeled objects. The conditional part of the model is based on a proposed conditional coupling layer. We demonstrate performance of the model for semi-supervised classification problem on different datasets. The model outperforms the baseline approach based on variational auto-encoders on MNIST dataset.