# Semi-Supervised Learning by Augmented Distribution Alignment

## TL;DR

This paper introduces Augmented Distribution Alignment, a semi-supervised learning method that reduces distribution mismatch between labeled and unlabeled data using adversarial training and data interpolation, improving performance on benchmark datasets.

## Contribution

It proposes a novel semi-supervised learning approach combining distribution alignment with adversarial training and data interpolation to address sampling bias.

## Key findings

- Effective on SVHN and CIFAR10 datasets.
- Reduces distribution mismatch between labeled and unlabeled data.
- Easily integrated into existing neural network frameworks.

## Abstract

In this work, we propose a simple yet effective semi-supervised learning approach called Augmented Distribution Alignment. We reveal that an essential sampling bias exists in semi-supervised learning due to the limited number of labeled samples, which often leads to a considerable empirical distribution mismatch between labeled data and unlabeled data. To this end, we propose to align the empirical distributions of labeled and unlabeled data to alleviate the bias. On one hand, we adopt an adversarial training strategy to minimize the distribution distance between labeled and unlabeled data as inspired by domain adaptation works. On the other hand, to deal with the small sample size issue of labeled data, we also propose a simple interpolation strategy to generate pseudo training samples. Those two strategies can be easily implemented into existing deep neural networks. We demonstrate the effectiveness of our proposed approach on the benchmark SVHN and CIFAR10 datasets. Our code is available at \url{https://github.com/qinenergy/adanet}.