MixupE: Understanding and improving Mixup from directional derivative perspective

Loading...
Thumbnail Image

Access rights

openAccess
publishedVersion

URL

Journal Title

Journal ISSN

Volume Title

A4 Artikkeli konferenssijulkaisussa

Date

2023-08

Major/Subject

Mcode

Degree programme

Language

en

Pages

Series

Proceedings of the 39th Conference on Uncertainty in Artificial Intelligence (UAI 2023), pp. 2597-2607, Proceedings of Machine Learning Research ; Volume 216

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.

Description

Keywords

Other note

Citation

Zou, Y, Verma, V, Mittal, S, Tang, W H, Pham, H, Kannala, J, Bengio, Y, Solin, A & Kawaguchi, K 2023, MixupE: Understanding and improving Mixup from directional derivative perspective . in Proceedings of the 39th Conference on Uncertainty in Artificial Intelligence (UAI 2023) . Proceedings of Machine Learning Research, vol. 216, JMLR, pp. 2597-2607, Conference on Uncertainty in Artificial Intelligence, Pittsburgh, Pennsylvania, United States, 31/07/2023 . < https://proceedings.mlr.press/v216/zou23a.html >