GeValDi: Generative Validation of Discriminative Models

Abstract

The evaluation of machine learning (ML) models is a core tenet of trustworthy use. Evaluation is typically done via a held-out dataset. However, such validation datasets often need to be large and are hard to procure; further, multiple models may perform equally well on such sets. To address these challenges, we offer GeValdi: an efficient method to validate discriminative classifiers by creating samples where such classifiers maximally differ. We demonstrate how such ``maximally different samples'' can be constructed via and leveraged to probe the failure mode of classifiers and offer a hierarchically-aware metric to further support fine-grained, comparative model evaluation.

Cite

Text

Palaniappan et al. "GeValDi: Generative Validation of Discriminative Models." ICLR 2023 Workshops: Trustworthy_ML, 2023.

Markdown

[Palaniappan et al. "GeValDi: Generative Validation of Discriminative Models." ICLR 2023 Workshops: Trustworthy_ML, 2023.](https://mlanthology.org/iclrw/2023/palaniappan2023iclrw-gevaldi/)

BibTeX

@inproceedings{palaniappan2023iclrw-gevaldi,
  title     = {{GeValDi: Generative Validation of Discriminative Models}},
  author    = {Palaniappan, Vivek and Ashman, Matthew and Collins, Katherine M. and Heo, Juyeon and Weller, Adrian and Bhatt, Umang},
  booktitle = {ICLR 2023 Workshops: Trustworthy_ML},
  year      = {2023},
  url       = {https://mlanthology.org/iclrw/2023/palaniappan2023iclrw-gevaldi/}
}