REDUCR: Robust Data Downsampling Using Class Priority Reweighting

William Bankes, George Hughes, Ilija Bogunovic, Zi Wang

Keywords: Class Robustness, Online Batch Selection, Robust Machine Learning, Training Efficiency, Data Downsampling, Class Imbalance

Abstract: Modern machine learning models are becoming increasingly expensive to train for real-world image and text classification tasks, where massive web-scale data is collected in a streaming fashion. To reduce the training cost, online batch selection techniques have been developed to choose the most informative datapoints. However, these techniques can suffer from poor worst-class generalization performance due to class imbalance and distributional shifts. This work introduces REDUCR, a robust and efficient data downsampling method that uses class priority reweighting. REDUCR reduces the training data while preserving worst-class generalization performance. REDUCR assigns priority weights to datapoints in a class-aware manner using an online learning algorithm. We demonstrate the data efficiency and robust performance of REDUCR on vision and text classification tasks. On web-scraped datasets with imbalanced class distributions, REDUCR achieves significant test accuracy boosts for the worst-performing class (but also on average), surpassing state-of-the-art methods by around 14%.

Submission Number: 14