Efficient Augmentation via Data Subsampling
This addresses storage and training cost issues for machine learning practitioners, but it is incremental as it builds on existing augmentation methods.
The paper tackles the inefficiency of data augmentation by showing that subsampling data points can achieve the same accuracy and invariance benefits as augmenting the entire dataset, with a 90% reduction in augmentation set size.
Data augmentation is commonly used to encode invariances in learning methods. However, this process is often performed in an inefficient manner, as artificial examples are created by applying a number of transformations to all points in the training set. The resulting explosion of the dataset size can be an issue in terms of storage and training costs, as well as in selecting and tuning the optimal set of transformations to apply. In this work, we demonstrate that it is possible to significantly reduce the number of data points included in data augmentation while realizing the same accuracy and invariance benefits of augmenting the entire dataset. We propose a novel set of subsampling policies, based on model influence and loss, that can achieve a 90% reduction in augmentation set size while maintaining the accuracy gains of standard data augmentation.