BriefGPT.xyz
Jul, 2020
URSABench：深度神经网络近似贝叶斯推断方法的全面基准测试
URSABench: Comprehensive Benchmarking of Approximate Bayesian Inference Methods for Deep Neural Networks
HTML
PDF
Meet P. Vadera, Adam D. Cobb, Brian Jalaian, Benjamin M. Marlin
TL;DR
本研究开发了一个基准测试套件URSABench，旨在对基于深度学习的分类任务的近似贝叶斯推理方法进行全面评估，解决其存在的不确定性、稳健性和计算规模可扩展性等问题。
Abstract
While
deep learning
methods continue to improve in predictive accuracy on a wide range of application domains, significant issues remain with other aspects of their performance including their ability to quantify
uncert
→