ARTICLE · INTELLIGENCE

战地情报 · 详情页

来自尧图项目组的一线实战观察与深度解析

RL-10-TD算法-ActorCritic02-离线算法02-赵:重要性采样【E_{X~p₀}[X]=∑p₀(x)x=∑p₁(x)·[p₀(x)/p₁(x)]x=E_{X~p₁}[f(X)]】

RL-10-TD算法-ActorCritic02-离线算法02-赵:重要性采样【E_{X~p₀}[X]=∑p₀(x)x=∑p₁(x)·[p₀(x)/p₁(x)]x=E_{X~p₁}[f(X)]】 2、重要性采样(Importance sampling)最核心的公式:EX∼p0[X]=∑xp0(x)x=∑xp1(x)p0(x)p1(x)x⏟f(x)=EX∼p1[f(X)]\color{red}{\mathbb{E}_{X \sim p_{0}}[X]
RELATED READING

延伸阅读

更多一线实战笔记与深度复盘,助您持续精进