Evaluating Bayesian deep learning for radio galaxy classification

Devina Mohan, Anna M M Scaife

Keywords: bayesian neural networks, variational inference, radio astronomy, distribution shift, uncertainty calibration

TL;DR: We present a domain-specific evaluation of Bayesian deep neural networks for radio astronomy

Abstract: The radio astronomy community is rapidly adopting deep learning techniques to deal with the huge data volumes expected from the next generation of radio observatories. Bayesian neural networks (BNNs) provide a principled way to model uncertainty in the predictions made by such deep learning models and will play an important role in extracting well-calibrated uncertainty estimates on their outputs. In this work, we evaluate the performance of different BNNs against the following criteria: predictive performance, uncertainty calibration and distribution-shift detection for the radio galaxy classification problem.

List Of Authors: Mohan, Devina and Scaife, Anna M M

Code Url: https://github.com/devinamhn/RadioGalaxies-BNNs

Submission Number: 750