|
| 1 | +site_name: 强化学习中的数学原理 |
| 2 | +site_url: https://wgyhhhh.github.io/Mathematical-Foundations-of-Reinforcement-Learning-Notes/ |
| 3 | +use_directory_urls: true |
| 4 | +repo_url: https://github.com/wgyhhhh/Mathematical-Foundations-of-Reinforcement-Learning-Notes |
| 5 | +repo_name: wgyhhhh/Mathematical-Foundations-of-Reinforcement-Learning-Notes |
| 6 | +site_author: Guangyu |
| 7 | +site_description: 强化学习中的数学原理 |
| 8 | + |
| 9 | +theme: |
| 10 | + icon: |
| 11 | + repo: fontawesome/brands/github-alt |
| 12 | + name: material |
| 13 | + palette: |
| 14 | + - media: "(prefers-color-scheme: light)" |
| 15 | + scheme: default |
| 16 | + primary: black |
| 17 | + accent: green |
| 18 | + toggle: |
| 19 | + icon: material/weather-sunny |
| 20 | + name: 明亮主题 |
| 21 | + - media: "(prefers-color-scheme: dark)" |
| 22 | + scheme: slate |
| 23 | + primary: black |
| 24 | + accent: yellow |
| 25 | + toggle: |
| 26 | + icon: material/weather-night |
| 27 | + name: 暗黑主题 |
| 28 | + logo: img/logo.jpg |
| 29 | + favicon: img/favicon-96x96.png |
| 30 | + custom_dir: docs/overrides |
| 31 | + |
| 32 | + features: |
| 33 | + - announce.dismiss #呈现可标记为由用户读取的临时公告,可以包含一个用于取消当前公告的按钮 |
| 34 | + # - navigation.instant |
| 35 | + #- header.autohide #自动隐藏 |
| 36 | + - navigation.tracking #地址栏中的 URL 将自动更新为在目录中突出显示的活动锚点 |
| 37 | + - navigation.tabs #顶级部分将呈现在上面视口标题下方的菜单层中,但在移动设备上保持原样 |
| 38 | + # - navigation.tabs.sticky #启用粘性选项卡后,导航选项卡将锁定在标题下方,并在向下滚动时始终保持可见 |
| 39 | + # - navigation.sections # 启用后会以分组形式常驻展开,不适合需要一级可折叠目录的场景 |
| 40 | + - navigation.top # 返回顶部的按钮 在上滑时出现 |
| 41 | + - navigation.footer #页脚将呈现在边栏中,但在移动设备上保持原样 |
| 42 | + - search.suggest # 搜索输入一些字母时推荐补全整个单词 |
| 43 | + - search.highlight # 搜索出的文章关键词加入高亮 |
| 44 | + - search.share #搜索分享按钮 |
| 45 | + - navigation.indexes #启用节索引页后,可以将文档直接附加到节 |
| 46 | + - content.tabs.link |
| 47 | + - content.tooltips |
| 48 | + - content.code.copy #代码复制 |
| 49 | + - content.action.edit |
| 50 | + - content.action.view |
| 51 | + - content.code.annotate |
| 52 | + language: zh # 一些提示性的文字会变成中文 |
| 53 | + |
| 54 | + |
| 55 | + |
| 56 | +extra: |
| 57 | + social: # 社交联系方式(页面右下角内容) |
| 58 | + - icon: fontawesome/brands/github |
| 59 | + link: https://github.com/wgyhhhh/ |
| 60 | + name: GitHub |
| 61 | + - icon: fontawesome/brands/bilibili |
| 62 | + link: https://space.bilibili.com/281217178 |
| 63 | + name: 哔哩哔哩 |
| 64 | + - icon: fontawesome/solid/paper-plane |
| 65 | + link: mailto:wgyhhh001@gmail.com |
| 66 | + name: 联系作者 |
| 67 | + version: |
| 68 | + provider: mike |
| 69 | + alias: true |
| 70 | + consent: |
| 71 | + actions: |
| 72 | + - accept |
| 73 | + - reject |
| 74 | + - manage |
| 75 | + title: Cookie 设置 |
| 76 | + description: >- |
| 77 | + 我们使用 cookies 来识别您的重复访问和偏好,以及衡量我们文档的有效性和用户是否找 |
| 78 | + 到他们正在搜索的内容。<br/> |
| 79 | + 在您的同意下,您将帮助我们改进我们的文档。<br/> |
| 80 | + (您稍后仍可以在网页左下角重新修改 cookies 设置) |
| 81 | + analytics: # 谷歌统计 |
| 82 | + provider: google |
| 83 | + property: G-C5GHN09YHC |
| 84 | + feedback: |
| 85 | + title: 该页面对你有帮助吗? |
| 86 | + ratings: |
| 87 | + - icon: material/emoticon-happy-outline |
| 88 | + name: 有帮助 |
| 89 | + data: 1 |
| 90 | + note: 感谢您的反馈! |
| 91 | + - icon: material/emoticon-neutral-outline |
| 92 | + name: 一般般 |
| 93 | + data: 0 |
| 94 | + note: 感谢您的反馈! |
| 95 | + - icon: material/emoticon-sad-outline |
| 96 | + name: 有待改进 |
| 97 | + data: -1 |
| 98 | + note: 感谢您的反馈! |
| 99 | +nav: |
| 100 | + - 强化学习中的数学原理笔记: |
| 101 | + - 主页: |
| 102 | + - 欢迎: 'index.md' |
| 103 | + - 第一版序言: 'Preface1.md' |
| 104 | + - 第二版序言: 'Preface2.md' |
| 105 | + - 第一章: |
| 106 | + - 介绍: 'Chapter-1/intro.md' |
| 107 | + - 1.1-网格世界: 'Chapter-1/1-1.md' |
| 108 | + - 1.2-状态和行动: 'Chapter-1/1-2.md' |
| 109 | + - 1.3-状态转移: 'Chapter-1/1-3.md' |
| 110 | + - 1.4-策略: 'Chapter-1/1-4.md' |
| 111 | + - 1.5-奖励: 'Chapter-1/1-5.md' |
| 112 | + - 1.6-轨迹、回报、回合: 'Chapter-1/1-6.md' |
| 113 | + - 1.7-马尔科夫决策过程: 'Chapter-1/1-7.md' |
| 114 | + - 1.8-总结: 'Chapter-1/1-8.md' |
| 115 | + - 第二章: |
| 116 | + - 介绍: 'Chapter-2/intro.md' |
| 117 | + - 2.1-为什么回报很重要?: 'Chapter-2/2-1.md' |
| 118 | + - 2.2-如何计算回报?: 'Chapter-2/2-2.md' |
| 119 | + - 2.3-状态值: 'Chapter-2/2-3.md' |
| 120 | + - 2.4-贝尔曼方程: 'Chapter-2/2-4.md' |
| 121 | + - 2.5-贝尔曼方程的例子: 'Chapter-2/2-5.md' |
| 122 | + - 2.6-贝尔曼方程的矩阵形式: 'Chapter-2/2-6.md' |
| 123 | + - 2.7-求解状态值: 'Chapter-2/2-7.md' |
| 124 | + - 2.8-行动值: 'Chapter-2/2-8.md' |
| 125 | + - 2.9-总结: 'Chapter-2/2-9.md' |
| 126 | + - 第三章: |
| 127 | + - 介绍: 'Chapter-3/intro.md' |
| 128 | + - 3.1-如何改进策略: 'Chapter-3/3-1.md' |
| 129 | + - 3.2-最优状态值和最优策略: 'Chapter-3/3-2.md' |
| 130 | + - 3.3-贝尔曼最优公式: 'Chapter-3/3-3.md' |
| 131 | + - 3.4-从贝尔曼最优公式中求解最优策略: 'Chapter-3/3-4.md' |
| 132 | + - 3.5-影响最优策略的因素: 'Chapter-3/3-5.md' |
| 133 | + - 3.6-总结: 'Chapter-3/3-6.md' |
| 134 | + - 第四章: |
| 135 | + - 介绍: 'Chapter-4/intro.md' |
| 136 | + - 4.1-值迭代: 'Chapter-4/4-1.md' |
| 137 | + - 4.2-策略迭代: 'Chapter-4/4-2.md' |
| 138 | + - 4.3-截断策略迭代: 'Chapter-4/4-3.md' |
| 139 | + - 4.4-总结: 'Chapter-4/4-4.md' |
| 140 | + - 第五章: |
| 141 | + - 介绍: 'Chapter-5/intro.md' |
| 142 | + - 5.1-启发示例:期望值估计: 'Chapter-5/5-1.md' |
| 143 | + - 5.2-MC Basic:最简单的基于蒙特卡洛的算法: 'Chapter-5/5-2.md' |
| 144 | + - 5.3-MC Exploring Starts算法: 'Chapter-5/5-3.md' |
| 145 | + - 5.4-MC-Greedy算法: 'Chapter-5/5-4.md' |
| 146 | + - 5.5-探索与利用:以Greedy策略为例: 'Chapter-5/5-5.md' |
| 147 | + - 5.6-总结: 'Chapter-5/5-6.md' |
| 148 | + - 第六章: |
| 149 | + - 介绍: 'Chapter-6/intro.md' |
| 150 | + - 6.1-启发示例:期望值估计: 'Chapter-6/6-1.md' |
| 151 | + - 6.2-罗宾斯-门罗算法: 'Chapter-6/6-2.md' |
| 152 | + - 6.3-Dvoretzky定理: 'Chapter-6/6-3.md' |
| 153 | + - 6.4-随机梯度下降: 'Chapter-6/6-4.md' |
| 154 | + - 6.5-总结: 'Chapter-6/6-5.md' |
| 155 | + - 第七章: |
| 156 | + - 介绍: 'Chapter-7/intro.md' |
| 157 | + - 7.1-状态值估计:时序差分算法: 'Chapter-7/7-1.md' |
| 158 | + - 7.2-行动值估计:Sarsa: 'Chapter-7/7-2.md' |
| 159 | + - 7.3-行动值估计:n步Sarsa: 'Chapter-7/7-3.md' |
| 160 | + - 7.4-最优行动值估计:Q-Learning: 'Chapter-7/7-4.md' |
| 161 | + - 7.5-时序差分算法的统一框架: 'Chapter-7/7-5.md' |
| 162 | + - 7.6-总结: 'Chapter-7/7-6.md' |
| 163 | + - 第八章: |
| 164 | + - 介绍: 'Chapter-8/intro.md' |
| 165 | + - 8.1-价值表示:从表格到函数: 'Chapter-8/8-1.md' |
| 166 | + - 8.2-基于值函数的时序差分算法:状态值估计: 'Chapter-8/8-2.md' |
| 167 | + - 8.3-基于值函数的时序差分算法:行动值估计: 'Chapter-8/8-3.md' |
| 168 | + - 8.4-深度Q-learning: 'Chapter-8/8-4.md' |
| 169 | + - 8.5-总结: 'Chapter-8/8-5.md' |
| 170 | + - 第九章: |
| 171 | + - 介绍: 'Chapter-9/intro.md' |
| 172 | + - 9.1-策略表示:从表格到函数: 'Chapter-9/9-1.md' |
| 173 | + - 9.2-目标函数:定义最优策略: 'Chapter-9/9-2.md' |
| 174 | + - 9.3-目标函数的梯度: 'Chapter-9/9-3.md' |
| 175 | + - 9.4-蒙特卡洛策略梯度: 'Chapter-9/9-4.md' |
| 176 | + - 9.5-总结: 'Chapter-9/9-5.md' |
| 177 | + - 第十章: |
| 178 | + - 介绍: 'Chapter-10/intro.md' |
| 179 | + - 10.1-最简单的演员-评论性方法: 'Chapter-10/10-1.md' |
| 180 | + - 10.2-优势演员-评论性方法: 'Chapter-10/10-2.md' |
| 181 | + - 10.3-异策略演员-评论性方法: 'Chapter-10/10-3.md' |
| 182 | + - 10.4-确定性演员-评论性方法: 'Chapter-10/10-4.md' |
| 183 | + - 10.5-总结: 'Chapter-10/10-5.md' |
| 184 | + - Box(证明): |
| 185 | + - 介绍: 'Box/intro.md' |
| 186 | + - 2.1:迭代解法的收敛性证明: 'Box/Box-2-1.md' |
| 187 | + - 7.1:TD算法的推导: 'Box/Box-7-1.md' |
| 188 | + - 7.4:期望Sarsa算法: 'Box/Box-7-4.md' |
| 189 | + - 附录: |
| 190 | + - 术语: 'Appendix/1.md' |
| 191 | + - 参考文献: 'References.md' |
| 192 | + - 核心算法实现: |
| 193 | + - 策略迭代与值迭代: 'python/code.ipynb' |
| 194 | + |
| 195 | +plugins: |
| 196 | + - search |
| 197 | + - mkdocs-jupyter |
| 198 | + |
| 199 | +markdown_extensions: #详见https://squidfunk.github.io/mkdocs-material/setup/extensions/python-markdown-extensions/ 和 https://squidfunk.github.io/mkdocs-material/setup/extensions/python-markdown/ |
| 200 | + - abbr |
| 201 | + - attr_list |
| 202 | + - admonition |
| 203 | + - def_list |
| 204 | + - footnotes |
| 205 | + - md_in_html |
| 206 | + - meta # 支持Markdown文件上方自定义标题标签等 |
| 207 | + - pymdownx.caret |
| 208 | + - pymdownx.betterem |
| 209 | + - pymdownx.critic |
| 210 | + - pymdownx.details |
| 211 | + - pymdownx.inlinehilite |
| 212 | + - pymdownx.keys |
| 213 | + - pymdownx.mark |
| 214 | + - pymdownx.snippets |
| 215 | + - pymdownx.smartsymbols |
| 216 | + - pymdownx.tilde |
| 217 | + - pymdownx.superfences: |
| 218 | + custom_fences: |
| 219 | + - name: mermaid |
| 220 | + class: mermaid |
| 221 | + format: |
| 222 | + !!python/name:pymdownx.superfences.fence_code_format # 代码块高亮插件 |
| 223 | + - pymdownx.arithmatex: # latex支持 |
| 224 | + generic: true |
| 225 | + - toc: |
| 226 | + permalink: true # 固定标题位置为当前位置 |
| 227 | + - pymdownx.highlight: # 代码块高亮 |
| 228 | + anchor_linenums: true |
| 229 | + linenums: true # 显示行号 |
| 230 | + # auto_title: true # 显示编程语言名称 |
| 231 | + - pymdownx.tabbed: |
| 232 | + alternate_style: true |
| 233 | + - pymdownx.tasklist: |
| 234 | + custom_checkbox: true |
| 235 | + |
| 236 | +extra_javascript: |
| 237 | + # - javascripts/extra.js #自定义javascript |
| 238 | + - https://Xiaokang2022.github.io/maliang/js/click-colorful.js # 鼠标点击效果 |
| 239 | + #- https://cdn.jsdelivr.net/npm/mathjax@2/MathJax.js #Latex支持 |
| 240 | + - https://polyfill.io/v3/polyfill.min.js?features=es6 #Latex支持 |
| 241 | + # - ckplayer/js/ckplayer.js #播放器配置 |
| 242 | + # - https://cdn.jsdelivr.net/npm/gitalk@latest/dist/gitalk.min.js #gitalk支持 |
| 243 | + - https://cdn.jsdelivr.net/npm/mermaid@10.0.2/dist/add-html-label-6e56ed67.min.js #忘了 |
| 244 | + - javascripts/extra.js |
| 245 | + - javascripts/mathjax.js #MathJax配置 |
| 246 | + - https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js # 单次加载,保留原始渲染风格 |
| 247 | + |
| 248 | + |
| 249 | +extra_css: |
| 250 | + - css/misc.css #自定义css |
| 251 | + - stylesheets/extra.css |
| 252 | +extra_templates: |
| 253 | +# - sitemap.xml |
| 254 | + |
| 255 | + |
0 commit comments