Data Augmentation for Traffic Classification accepted at PAM 2024

Our abstract titled “Data Augmentation for Traffic Classification” will be presented at PAM 2024

Congrats to Wang Chao and other Co-authors.

Abstract: Abstract. Data Augmentation (DA)—enriching training data by adding synthetic samples—is a technique widely adopted in the Computer Vision (CV) and Natural Language Processing (NLP) domains to improve models performance. Yet, DA has struggled to gain traction in networking contexts, particularly in Traffic Classification (TC) tasks. In this work, we fulfill this gap by benchmarking 18 augmentation functions applied to 3 TC datasets across a variety of conditions. Our results (i) show that DA can reap benefits previously unexplored with (ii) augmentations acting on sequence order and masking being a better suit for TC and (iii) provide hints about why augmentations have positive or negative effects based on simple latent space analysis.