Confidence Scoring Using Whitebox Meta-models with Linear Classifier Probes
Tongfei Chen, Jiri Navratil, Vijay Iyengar, Karthikeyan Shanmugam
Abstract
We propose a novel confidence scoring mechanism for deep neural networks based on a two-model paradigm involving a base model and a meta-model. The confidence score is learned by the meta-model observing the base model succeeding/failing at its task. As features to the meta-model, we investigate linear classifier probes inserted between the various layers of the base model. Our experiments demonstrate that this approach outperforms multiple baselines in a filtering task, i.e., task of rejecting samples with low confidence. Experimental results are presented using CIFAR-10 and CIFAR-100 dataset with and without added noise. We discuss the importance of confidence scoring to bridge the gap between experimental and real-world applications.
BibTeX
@InProceedings{pmlr-v89-chen19c,
title = {Confidence Scoring Using Whitebox Meta-models with Linear Classifier Probes},
author = {Chen, Tongfei and Navratil, Jiri and Iyengar, Vijay and Shanmugam, Karthikeyan},
booktitle = {Proceedings of the Twenty-Second International Conference on Artificial Intelligence and Statistics},
pages = {1467--1475},
year = {2019},
editor = {Chaudhuri, Kamalika and Sugiyama, Masashi},
volume = {89},
series = {Proceedings of Machine Learning Research},
month = {16--18 Apr},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v89/chen19c/chen19c.pdf},
url = {https://proceedings.mlr.press/v89/chen19c.html},
abstract = {We propose a novel confidence scoring mechanism for deep neural networks based on a two-model paradigm involving a base model and a meta-model. The confidence score is learned by the meta-model observing the base model succeeding/failing at its task. As features to the meta-model, we investigate linear classifier probes inserted between the various layers of the base model. Our experiments demonstrate that this approach outperforms multiple baselines in a filtering task, i.e., task of rejecting samples with low confidence. Experimental results are presented using CIFAR-10 and CIFAR-100 dataset with and without added noise. We discuss the importance of confidence scoring to bridge the gap between experimental and real-world applications.}
}