Sampling-based Bayesian Inference with gradient uncertainty
This addresses uncertainty estimation for safety-critical applications, but appears incremental as it builds on existing Bayesian inference methods.
The paper tackles the problem of neural networks being overconfident in predictions, which is harmful in safety-critical applications, by incorporating gradient uncertainty into posterior sampling to efficiently estimate predictive uncertainty, showing effectiveness on MNIST and notMNIST datasets.
Deep neural networks(NNs) have achieved impressive performance, often exceed human performance on many computer vision tasks. However, one of the most challenging issues that still remains is that NNs are overconfident in their predictions, which can be very harmful when this arises in safety critical applications. In this paper, we show that predictive uncertainty can be efficiently estimated when we incorporate the concept of gradients uncertainty into posterior sampling. The proposed method is tested on two different datasets, MNIST for in-distribution confusing examples and notMNIST for out-of-distribution data. We show that our method is able to efficiently represent predictive uncertainty on both datasets.