BriefGPT.xyz
Oct, 2019
带矩阵值核的斯坦变分梯度下降
Stein Variational Gradient Descent With Matrix-Valued Kernels
HTML
PDF
Dilin Wang, Ziyang Tang, Chandrajit Bajaj, Qiang Liu
TL;DR
本文提出了一种新颖的基于矩阵的 Stein 变分梯度下降算法，通过利用 Hessian 矩阵和 Fisher 信息矩阵等预处理矩阵来加速粒子的探索，从而实现了更加高效的近似推断，并在实验中证明其性能优于其他基线方法。
Abstract
stein variational gradient descent
(SVGD) is a
particle-based inference
algorithm that leverages gradient information for efficient approximate inference. In this work, we enhance SVGD by leveraging
→