MixupE: Understanding and Improving Mixup from Directional Derivative Perspective
Abstract
Mixup is a popular data augmentation technique for training deep neural networks where additional samples are generated by linearly interpolating pairs of inputs and their labels. This technique is known to improve the generalization performance in many learning paradigms and applications. In this work, we first analyze Mixup and show that it implicitly regularizes infinitely many directional derivatives of all orders. Based on this new insight, we propose an improved version of Mixup, theoretically justified to deliver better generalization performance than the vanilla Mixup. To demonstrate the effectiveness of the proposed method, we conduct experiments across various domains such as images, tabular data, speech, and graphs. Our results show that the proposed method improves Mixup across multiple datasets using a variety of architectures, for instance, exhibiting an improvement over Mixup by 0.8% in ImageNet top-1 accuracy.
Cite
@article{arxiv.2212.13381,
title = {MixupE: Understanding and Improving Mixup from Directional Derivative Perspective},
author = {Yingtian Zou and Vikas Verma and Sarthak Mittal and Wai Hoh Tang and Hieu Pham and Juho Kannala and Yoshua Bengio and Arno Solin and Kenji Kawaguchi},
journal= {arXiv preprint arXiv:2212.13381},
year = {2023}
}
Comments
16 pages, Best Student Paper Award at UAI 2023