A Comparison of the Delta Method and the Bootstrap in Deep Learning Classification
Machine Learning
2021-07-06 v1 Machine Learning
Abstract
We validate the recently introduced deep learning classification adapted Delta method by a comparison with the classical Bootstrap. We show that there is a strong linear relationship between the quantified predictive epistemic uncertainty levels obtained from the two methods when applied on two LeNet-based neural network classifiers using the MNIST and CIFAR-10 datasets. Furthermore, we demonstrate that the Delta method offers a five times computation time reduction compared to the Bootstrap.
Keywords
Cite
@article{arxiv.2107.01606,
title = {A Comparison of the Delta Method and the Bootstrap in Deep Learning Classification},
author = {Geir K. Nilsen and Antonella Z. Munthe-Kaas and Hans J. Skaug and Morten Brun},
journal= {arXiv preprint arXiv:2107.01606},
year = {2021}
}