Learning to Augment Influential Data

Data augmentation is a technique to reduce overfitting and to improve generalization by increasing the number of labeled data samples by performing label preserving transformations; however, it is currently conducted in a trial and error manner. A composition of predefined transformations, such as rotation, scaling and cropping, is performed on training samples, and its effect on performance over test samples can only be empirically evaluated and cannot be predicted. This paper considers an influence function which predicts how generalization is affected by a particular augmented training sample in terms of validation loss. The influence function provides an approximation of the change in validation loss without comparing the performance which includes and excludes the sample in the training process. A differentiable augmentation model that generalizes the conventional composition of predefined transformations is also proposed. The differentiable augmentation model and reformulation of the influence function allow the parameters of the augmented model to be directly updated by backpropagation to minimize the validation loss. The experimental results show that the proposed method provides better generalization over conventional data augmentation methods.