Retrospective Uncertainties for Deep Models using Vine Copulas
Natasa Tagasovska, Firat Ozdemir, Axel Brando
Abstract
Despite the major progress of deep models as learning machines, uncertainty estimation remains a major challenge. Existing solutions rely on modified loss functions or architectural changes. We propose to compensate for the lack of built-in uncertainty estimates by supplementing any network, retrospectively, with a subsequent vine copula model, in an overall compound we call Vine-Copula Neural Network (VCNN). Through synthetic and real-data experiments, we show that VCNNs could be task (regression/classification) and architecture (recurrent, fully connected) agnostic while providing reliable and better-calibrated uncertainty estimates, comparable to state-of-the-art built-in uncertainty solutions.
BibTeX
@InProceedings{pmlr-v206-tagasovska23a,
title = {Retrospective Uncertainties for Deep Models using Vine Copulas},
author = {Tagasovska, Natasa and Ozdemir, Firat and Brando, Axel},
booktitle = {Proceedings of The 26th International Conference on Artificial Intelligence and Statistics},
pages = {7528--7539},
year = {2023},
editor = {Ruiz, Francisco and Dy, Jennifer and van de Meent, Jan-Willem},
volume = {206},
series = {Proceedings of Machine Learning Research},
month = {25--27 Apr},
publisher = {PMLR},
pdf = {https://proceedings.mlr.press/v206/tagasovska23a/tagasovska23a.pdf},
url = {https://proceedings.mlr.press/v206/tagasovska23a.html},
abstract = {Despite the major progress of deep models as learning machines, uncertainty estimation remains a major challenge. Existing solutions rely on modified loss functions or architectural changes. We propose to compensate for the lack of built-in uncertainty estimates by supplementing any network, retrospectively, with a subsequent vine copula model, in an overall compound we call Vine-Copula Neural Network (VCNN). Through synthetic and real-data experiments, we show that VCNNs could be task (regression/classification) and architecture (recurrent, fully connected) agnostic while providing reliable and better-calibrated uncertainty estimates, comparable to state-of-the-art built-in uncertainty solutions.}
}