lh15123as的博客

@[TOC]强化学习 Chapter 4. Dynamic Programming (4.1)v(s)=max⁡aE[Rt+1+γv(St+1)∣St=s,At=a] =max⁡a∑s′,rp(s′,r∣s,a)[r+γv∗(s′)] \begin{aligned} v_(s)& = \max a\mathbb{E}[R{t+1}+\gamma v_(S_{t+1}) |...

文章目录5.1 Monte Carlo Prediction5.2 Monte Carlo Estimation of Action Values5.3 Monte Carlo Control5.4 Monte Carlo Control without Exploring Starts5.5 Off-policy Prediction via Importance Sampling5.6 Inc...

文章目录6.1 TD Prediction 6.1 TD Prediction

文章目录7.1 n-step TD Prediction7.2 n-step Sarsa7.3 n-step Off-policy Learning by Importance Sampling 7.1 n-step TD Prediction 输入：策略 :π\piπ 算法参数：步长 α∈(0,1]\alpha \in (0,1]α∈(0,1]，正整数 nnn 对 s∈Ss \in \math...