Uncertainty Quantification in CNN Through the Bootstrap of Convex Neural Networks

TL;DR

This paper introduces a bootstrap-based framework for uncertainty quantification in CNNs using convex neural networks, offering theoretical guarantees and lower computational costs.

Contribution

It proposes a novel, theoretically consistent bootstrap method for CNN uncertainty estimation that is computationally efficient and adaptable via transfer learning.

Findings

Outperforms baseline CNNs and state-of-the-art methods in uncertainty estimation.

Provides theoretical guarantees for the quality of uncertainty estimates.

Reduces computational load compared to existing approaches.

Abstract

Despite the popularity of Convolutional Neural Networks (CNN), the problem of uncertainty quantification (UQ) of CNN has been largely overlooked. Lack of efficient UQ tools severely limits the application of CNN in certain areas, such as medicine, where prediction uncertainty is critically important. Among the few existing UQ approaches that have been proposed for deep learning, none of them has theoretical consistency that can guarantee the uncertainty quality. To address this issue, we propose a novel bootstrap based framework for the estimation of prediction uncertainty. The inference procedure we use relies on convexified neural networks to establish the theoretical consistency of bootstrap. Our approach has a significantly less computational load than its competitors, as it relies on warm-starts at each bootstrap that avoids refitting the model from scratch. We further explore a novel transfer learning method so our framework can work on arbitrary neural networks. We experimentally demonstrate our approach has a much better performance compared to other baseline CNNs and state-of-the-art methods on various image datasets.