Virtual Adversarial Training: A Regularization Method for Supervised and Semi-Supervised Learning
This addresses the need for effective regularization in machine learning, particularly for semi-supervised scenarios where labeled data is scarce, though it is incremental as it builds on adversarial training concepts.
The paper tackles the problem of improving model robustness and performance in supervised and semi-supervised learning by proposing virtual adversarial training (VAT), a regularization method based on local smoothness of the conditional label distribution. It achieves state-of-the-art results on SVHN and CIFAR-10 datasets for semi-supervised learning.
We propose a new regularization method based on virtual adversarial loss: a new measure of local smoothness of the conditional label distribution given input. Virtual adversarial loss is defined as the robustness of the conditional label distribution around each input data point against local perturbation. Unlike adversarial training, our method defines the adversarial direction without label information and is hence applicable to semi-supervised learning. Because the directions in which we smooth the model are only "virtually" adversarial, we call our method virtual adversarial training (VAT). The computational cost of VAT is relatively low. For neural networks, the approximated gradient of virtual adversarial loss can be computed with no more than two pairs of forward- and back-propagations. In our experiments, we applied VAT to supervised and semi-supervised learning tasks on multiple benchmark datasets. With a simple enhancement of the algorithm based on the entropy minimization principle, our VAT achieves state-of-the-art performance for semi-supervised learning tasks on SVHN and CIFAR-10.