Files
easy-rl/docs/errata.md
2022-04-01 09:20:23 +08:00

43 lines
1.8 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# 纸质版勘误表
如何使用勘误?首先找到你的书的印次,接下来对着下表索引印次,该印次之后所有的勘误都是你的书中所要注意的勘误,印次前的所有勘误在当印次和之后印次均已印刷修正。
## 第1版第1次印刷2022.03
* 35页的图2.2 和 41页的图2.5a替换成下图
![](res/Markov_chain.png ':size=450')
* 47页2.3.5节的第3行称为备份图backup diagram → 称为备份图backup diagram或回溯图
* 76页式(3.1) 中 $G$ 和 $r$ 后面的数字改为下标,即
$$
\begin{array}{l}
G_{13}=0 \\
G_{12}=r_{13}+\gamma G_{13}=-1+0.6 \times 0=-1 \\
G_{11}=r_{12}+\gamma G_{12}=-1+0.6 \times(-1)=-1.6 \\
G_{10}=r_{11}+\gamma G_{11}=-1+0.6 \times(-1.6)=-1.96 \\
G_9=r_{10}+\gamma G_{10}=-1+0.6 \times(-1.96)=-2.176 \approx-2.18 \\
G_8=r_9+\gamma G_9=-1+0.6 \times(-2.176)=-2.3056 \approx-2.3
\end{array}
$$
* 149页式(6.15) 改为
$$
\begin{aligned}
V^{\pi}(s) &\le Q^{\pi}(s,\pi'(s)) \\
&=E\left[r_{t}+V^{\pi}\left(s_{t+1}\right) | s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right]\\
&\le E\left[r_{t}+Q^{\pi}\left(s_{t+1}, \pi^{\prime}\left(s_{t+1}\right)\right) | s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right] \\
&=E\left[r_{t}+r_{t+1}+V^{\pi}\left(s_{t+2}\right) |s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right] \\
& \le E\left[r_{t}+r_{t+1}+Q^{\pi}\left(s_{t+2},\pi'(s_{t+2}\right) | s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right] \\
& = E\left[r_{t}+r_{t+1}+r_{t+2}+V^{\pi}\left(s_{t+3}\right) |s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right] \\
& \le \cdots\\
& \le E\left[r_{t}+r_{t+1}+r_{t+2}+\cdots | s_{t}=s, a_{t}=\pi^{\prime}\left(s_{t}\right)\right] \\
& = V^{\pi'}(s)
\end{aligned}
$$
* 229页第2行很强的序列 → 很长的序列