BriefGPT.xyz
Mar, 2018
实践中的自然梯度：高斯过程模型中的非共轭变分推断
Natural Gradients in Practice: Non-Conjugate Variational Inference in Gaussian Process Models
HTML
PDF
Hugh Salimbeni, Stefanos Eleftheriadis, James Hensman
TL;DR
本文探讨了利用自然梯度方法在非共轭随机模型环境下的超参数学习，结果显示自然梯度方法可显著提高性能和效率并已被集成于GPflow软件包中。
Abstract
The
natural gradient method
has been used effectively in conjugate Gaussian process models, but the non-conjugate case has been largely unexplored. We examine how natural gradients can be used in non-conjugate
stochasti
→