Combining Statistical Depth and Fermat Distance for Uncertainty Quantification

Abstract

We measure the out-of-domain uncertainty in the prediction of Neural Networks using a statistical notion called "Lens Depth'' (LD) combined with Fermat Distance, which is able to capture precisely the "depth'' of a point with respect to a distribution in feature space, without any distributional assumption. Our method also has no trainable parameter. The method is applied directly in the feature space at test time and does not intervene in training process. As such, it does not impact the performance of the original model. The proposed method gives excellent qualitative results on toy datasets and can give competitive or better uncertainty estimation on standard deep learning datasets compared to strong baseline methods.

PDF NeurIPS OpenReview Semantic Scholar

Cite

Text

Nguyen et al. "Combining Statistical Depth and Fermat Distance for Uncertainty Quantification." Neural Information Processing Systems, 2024. doi:10.52202/079017-0951

Markdown

[Nguyen et al. "Combining Statistical Depth and Fermat Distance for Uncertainty Quantification." Neural Information Processing Systems, 2024.](https://mlanthology.org/neurips/2024/nguyen2024neurips-combining/) doi:10.52202/079017-0951

BibTeX

@inproceedings{nguyen2024neurips-combining,
  title     = {{Combining Statistical Depth and Fermat Distance for Uncertainty Quantification}},
  author    = {Nguyen, Hai-Vy and Gamboa, Fabrice and Chhaibi, Reda and Zhang, Sixin and Gratton, Serge and Giaccone, Thierry},
  booktitle = {Neural Information Processing Systems},
  year      = {2024},
  doi       = {10.52202/079017-0951},
  url       = {https://mlanthology.org/neurips/2024/nguyen2024neurips-combining/}
}