Improve Uncertainty Estimation for Unknown Classes in Bayesian Neural Networks with Semi-Supervised /One Set Classification

TL;DR

This paper enhances Bayesian Neural Networks' ability to estimate uncertainty for unknown classes by integrating set classification, addressing safety-critical issues like misclassification of unseen data.

Contribution

The paper introduces a simple method combining BNN with set classification to better detect unknown classes, improving uncertainty estimation.

Findings

Improved detection of unknown classes in BNNs.

Effective on MNIST, notMNIST, and FMNIST datasets.

Addresses safety concerns in critical applications.

Abstract

Although deep neural network (DNN) has achieved many state-of-the-art results, estimating the uncertainty presented in the DNN model and the data is a challenging task. Problems related to uncertainty such as classifying unknown classes (class which does not appear in the training data) data as known class with high confidence, is critically concerned in the safety domain area (e.g, autonomous driving, medical diagnosis). In this paper, we show that applying current Bayesian Neural Network (BNN) techniques alone does not effectively capture the uncertainty. To tackle this problem, we introduce a simple way to improve the BNN by using one class classification (in this paper, we use the term "set classification" instead). We empirically show the result of our method on an experiment which involves three datasets: MNIST, notMNIST and FMNIST.