ThinkNotClear
  • Home
  • Categories
  • Tags
  • Archives

Score Matching: Estimation of non-normalized statistical models by score matching

目录

Table of Contents
  • 概述
目录

概述

$p_{\theta}(x) = \frac{e^{-E_\theta(x)}}{Z(\theta)}$, $Z(\theta) = \int_x e^{-E_\theta(x)}$配分函数 是对所有 x 的积分, 不可计算, 但是对x 来说是个常数。

所以 最大对数似然估计 MLE 对 $\theta$求导更新 需要 MCMC采样, 很慢

但 $p_{\theta}(x)$


Published

Sep 19, 2022

Category

paper

Tags

  • generator 8
  • paper 9
  • score matching 3
  • Powered by Pelican. Theme: Elegant by Talha Mansoor