Learning by Association — A Versatile Semi-Supervised Training Method for Neural Networks

Citations Over TimeTop 10% of 2017 papers

Abstract

In many real-world scenarios, labeled data for a specific machine learning task is costly to obtain. Semi-supervised training methods make use of abundantly available unlabeled data and a smaller number of labeled examples. We propose a new framework for semi-supervised training of deep neural networks inspired by learning in humans. Associations are made from embeddings of labeled samples to those of unlabeled ones and back. The optimization schedule encourages correct association cycles that end up at the same class from which the association was started and penalizes wrong associations ending at a different class. The implementation is easy to use and can be added to any existing end-to-end training setup. We demonstrate the capabilities of learning by association on several data sets and show that it can improve performance on classification tasks tremendously by making use of additionally available unlabeled data. In particular, for cases with few labeled data, our training scheme outperforms the current state of the art on SVHN.

Related Papers

• → Semi-supervised learning: a brief review(2018)232 cited
• → The Use of Unlabeled Data Versus Labeled Data for Stopping Active Learning for Text Classification(2019)17 cited
• → TWSVM for Unsupervised and Semi-supervised Learning(2016)4 cited
• → Estimate Unlabeled-Data-Distribution for Semi-supervised PU Learning(2012)1 cited
• → Recommender system designed using an ensemble approach to semi-supervised learning(2022)1 cited