Improved Text Classification via Test-Time Augmentation
This work addresses the problem of improving text classification performance for NLP practitioners, but it is incremental as it adapts an established technique from image classification to NLP.
The paper tackled the limited adoption of test-time augmentation (TTA) in NLP by developing augmentation policies for language models, resulting in consistent accuracy improvements over state-of-the-art approaches in binary classification tasks.
Test-time augmentation -- the aggregation of predictions across transformed examples of test inputs -- is an established technique to improve the performance of image classification models. Importantly, TTA can be used to improve model performance post-hoc, without additional training. Although test-time augmentation (TTA) can be applied to any data modality, it has seen limited adoption in NLP due in part to the difficulty of identifying label-preserving transformations. In this paper, we present augmentation policies that yield significant accuracy improvements with language models. A key finding is that augmentation policy design -- for instance, the number of samples generated from a single, non-deterministic augmentation -- has a considerable impact on the benefit of TTA. Experiments across a binary classification task and dataset show that test-time augmentation can deliver consistent improvements over current state-of-the-art approaches.