BriefGPT.xyz
May, 2017
核隐式变分推断
Implicit Variational Inference with Kernel Density Ratio Fitting
HTML
PDF
Jiaxin Shi, Shengyang Sun, Jun Zhu
TL;DR
本文介绍一种名为核式隐式变分推断的新方法，可以成功地将隐式变分推断应用于贝叶斯神经网络，并在回归和分类任务上展现了有前途的结果。
Abstract
Recent progress in
variational inference
has paid much attention to the flexibility of variational posteriors. Work has been done to use
implicit distributions
, i.e., distributions without tractable likelihoods a
→