├── README.md ├── slides ├── 2 │ ├── topics │ │ ├── swarm-files │ │ │ ├── lzq_0.png │ │ │ ├── lzq_1.png │ │ │ ├── lzq_2.png │ │ │ ├── lzq_3.png │ │ │ ├── lzq_4.png │ │ │ └── lzq_5.png │ │ ├── forage.tex │ │ ├── evolution.tex │ │ ├── cn.tex │ │ ├── immunity.tex │ │ ├── ann.tex │ │ └── swarm.tex │ └── 2.tex ├── 0-files │ ├── ant.png │ └── qzn.png ├── 3.tex ├── 1.tex ├── 4.tex └── 0.tex ├── references └── 0.tex ├── .gitignore └── LICENSE /README.md: -------------------------------------------------------------------------------- 1 | # OrganicIntelligentCourseware 2 | Courseware of organic intelligence 3 | -------------------------------------------------------------------------------- /slides/0-files/ant.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/0-files/ant.png -------------------------------------------------------------------------------- /slides/0-files/qzn.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/0-files/qzn.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_0.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_0.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_1.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_1.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_2.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_2.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_3.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_3.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_4.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_4.png -------------------------------------------------------------------------------- /slides/2/topics/swarm-files/lzq_5.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/Zhejiang-University-GKC/OrganicIntelligentCourseware/HEAD/slides/2/topics/swarm-files/lzq_5.png -------------------------------------------------------------------------------- /references/0.tex: -------------------------------------------------------------------------------- 1 | 2 | \frame{\frametitle{课程报告} 3 | \begin{enumerate} 4 | \item 上课时各小组的幻灯或小报告 5 | \item 每位同学整理一份报告 6 | \begin{itemize} 7 | \item 对已有的一个方法的深入阐述,有余力的同学可编程实验并评价 8 | \item 探索新的方法 9 | \item 针对科研中的问题,提出计算智能方法并试验 10 | \end{itemize} 11 | \item 协作网址或报告发邮件给 yxxinyuan@zju.edu.cn 12 | \end{enumerate} 13 | } 14 | 15 | \frame{\frametitle{报告形式} 16 | \begin{minipage}[t]{\textwidth} 17 | 幻灯和报告可充分应用图形摘要、漫画等趣味形式,形成既科普,又专业的作品。 18 | \end{minipage} 19 | } 20 | -------------------------------------------------------------------------------- /slides/2/topics/forage.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{觅食算法} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{觅食算法}}} 9 | } 10 | 11 | \frame{\frametitle{思想} 12 | 13 | {\CJKfamily{zhfs} 14 | 来自生物\textbf{高效觅食}的启发。包括细菌、肉食动物等。 15 | 另外还来自布谷鸟寄生算法的启发。 16 | } 17 | 18 | ~ 19 | 20 | 仍然是从一个初始的点出发在函数空间中搜索解。所有样本都参与计算。 21 | } 22 | 23 | \frame{\frametitle{思想} 24 | \begin{itemize} 25 | \item<1-> 布朗运动 26 | \item<2-> 随机游走 27 | \item<3-> 随机微分方程和伊藤过程 28 | \item<4-> 轻尾和重尾分布 29 | \end{itemize} 30 | } 31 | 32 | \frame{\frametitle{算法} 33 | \begin{itemize} 34 | \item<1-> 莱维过程(L\'evy Process) 35 | \item<2-> 分解定理(布朗+泊松+平方可积离散鞅,可类比PID) 36 | \end{itemize} 37 | } 38 | 39 | %end 40 | -------------------------------------------------------------------------------- /slides/2/topics/evolution.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{进化算法} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{进化算法}}} 9 | } 10 | 11 | \frame{\frametitle{定义} 12 | 又叫演化计算,是模拟自然界中的生物的演化过程产生的一种群体导向的随机搜索技术和方法。 13 | 14 | ~ 15 | 16 | 是一种通用的问题求解方法,具有自组织、自适应、自学习性和本质并行性等特点, 17 | 不受搜索空间限制性条件的约束,也不需要其它辅助信息。 18 | } 19 | 20 | \frame{\frametitle{思想} 21 | 其算法是受生物进化过程中“优胜劣汰”的自然选择机制和遗传信息的传递规律的影响, 22 | 通过程序迭代模拟这一过程,把要解决的问题看作环境,在一些可能的解组成的种群中, 23 | 通过自然演化寻求最优解。 24 | } 25 | 26 | \frame{\frametitle{种类} 27 | \begin{itemize} 28 | \item<1-> 遗传算法 29 | \item<2-> 演化策略,修改参数 30 | \item<3-> 演化规划,自适应响应 31 | \item<4-> 遗传程序设计,改变分层节点和结构链接关系 32 | \item<5-> 多种群协同进化,竞争和合作的动力学系统 33 | \item<6-> 差分进化算法,个体间竞争和合作的系统 34 | \end{itemize} 35 | } 36 | 37 | %end 38 | -------------------------------------------------------------------------------- /slides/2/topics/cn.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{复杂网络} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{复杂网络}}} 9 | } 10 | 11 | \frame{\frametitle{定义} 12 | 13 | 具有自组织、自相似、吸引子、小世界、无标度中部分或全部性质的网络称为复杂网络。 14 | } 15 | 16 | \frame{\frametitle{思想} 17 | 18 | 互联网拓扑、社交网络、脑神经网络、基因调控网络、蛋白质调控网络、生态系统、地球演化、N体作用。 19 | 可类比去中心化思想。 20 | } 21 | 22 | \frame{\frametitle{思想} 23 | \begin{itemize} 24 | \item<1-> 结构复杂:节点数巨大,结构具有多种不同特征。 25 | \item<2-> 网络进化:表现在节点或连接的产生与消失,网络结构不断变化。 26 | \item<3-> 连接多样性:节点之间的连接权重存在差异,且有可能存在方向性。 27 | \item<4-> 动力学复杂性:节点集可能属于非线性动力学系统,节点状态随时间发生复杂变化。 28 | \item<5-> 节点多样性:节点和连接可代表任何事物。 29 | \item<6-> 多重复杂性融合:多重复杂性相互影响,结果难以预测。 30 | \end{itemize} 31 | } 32 | 33 | \frame{\frametitle{思想} 34 | 网络的几何性质,网络的形成机制,网络演化的统计规律,网络上的模型性质, 35 | 网络的结构稳定性,网络的演化动力学机制。 36 | } 37 | 38 | \frame{\frametitle{种类} 39 | \begin{itemize} 40 | \item<1-> 马尔科夫场及条件随机场 41 | \item<2-> 概率图和贝叶斯网络 42 | \item<3-> 社交网络 43 | \item<4-> 视觉听觉网络 44 | \item<5-> 大脑生理病理网络 45 | \item<6-> 组分调控网络 46 | \end{itemize} 47 | } 48 | 49 | \frame{\frametitle{种类} 50 | \begin{itemize} 51 | \item<1-> 传染病网络 52 | \item<2-> 生态系统网络(研究方法:微分方程组,混沌分形,元胞自动机) 53 | \item<3-> 网络安全攻防 54 | \item<4-> 天文N体系统 55 | \end{itemize} 56 | } 57 | 58 | %end 59 | -------------------------------------------------------------------------------- /slides/2/topics/immunity.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{免疫算法} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{免疫算法}}} 9 | } 10 | 11 | \frame{\frametitle{定义} 12 | 13 | \begin{block}{定义} 14 | 一种具有生成+检测 (generate and test) 的迭代过程的搜索算法。 15 | \end{block} 16 | } 17 | 18 | \frame{\frametitle{思想} 19 | 生物免疫系统是一个分布式、自组织和具有动态平衡能力的自适应复杂系统。 20 | 它对外界入侵的抗原,可由分布全身的不同种类的淋巴细胞产生相应的抗体, 21 | 其目标是尽可能保证整个生物系统的基本生理功能得到正常运转。 22 | 23 | ~ 24 | 25 | 具有较强模式分类能力,尤其对多模态问题的分析、处理和求解表现出较高的智能性和鲁棒性。 26 | } 27 | 28 | \frame{\frametitle{思想} 29 | \begin{itemize} 30 | \item<1-> 抗原和抗体 31 | \item<2-> 免疫疫苗 32 | \item<3-> 免疫算子 33 | \item<4-> 免疫调节 34 | \item<5-> 免疫记忆 35 | \item<6-> 抗原识别 36 | \end{itemize} 37 | } 38 | 39 | \frame{\frametitle{种类} 40 | \begin{itemize} 41 | \item<1-> 一般免疫算法 42 | \item<2-> 阴性选择算法 43 | \item<3-> 克隆选择算法 44 | \item<4-> 免疫网络学说与人工免疫网络模型 45 | \item<5-> B细胞网络算法 46 | \end{itemize} 47 | } 48 | 49 | \frame{\frametitle{种类} 50 | \begin{itemize} 51 | \item<1-> 危险理念(树突细胞,组织系统,TLRA) 52 | \item<2-> Cohen协同模型 53 | \item<3-> 免疫遗传算法 54 | \item<4-> 混合免疫算法 55 | \end{itemize} 56 | } 57 | 58 | \frame{\frametitle{发展} 59 | 在基于马尔科夫链的收敛性分析和非线性动力学模型等方面, 60 | 对免疫优化算法的非线性随机分析可能是未来研究的难点之一。 61 | 62 | ~ 63 | 64 | 神经网络、内分泌及免疫这三大调节系统相互联系、相互补充和配合、相互制约的机理 65 | 为基于人工免疫系统的智能综合集成提供了生物学基础, 66 | 网络和智能成为免疫算法发展的不可缺少的特征,也是其重要应用领域。 67 | 68 | ~ 69 | 70 | 免疫算法能增强系统的鲁棒性,在网络、智能系统和鲁棒系统中的应用。 71 | } 72 | 73 | \frame{\frametitle{科普} 74 | 人体防御病毒的免疫力究竟是什么? 75 | 76 | ~ 77 | 78 | 参见\href{http://news.sciencenet.cn/htmlnews/2020/2/436076.shtm}{网页} 79 | } 80 | 81 | %end 82 | -------------------------------------------------------------------------------- /slides/2/2.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \documentclass[11pt]{beamer} 6 | \usepackage{beamerthemesplit} 7 | \usepackage{algorithm, algorithmic} 8 | \usepackage{tikz} 9 | %url 10 | \usepackage{hyperref} 11 | \hypersetup{colorlinks=true, urlcolor=blue} 12 | %animate 13 | \usepackage{graphicx} 14 | \usepackage{animate} 15 | %flowchart 16 | \usepackage{mathpazo} 17 | \usetikzlibrary{shapes.geometric, arrows} 18 | \usepackage{flowchart} 19 | %xeCJK 20 | \usepackage[BoldFont,SlantFont,CJKchecksingle]{xeCJK} 21 | \setCJKmainfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 22 | \setCJKsansfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 23 | \setCJKmonofont[ItalicFont={Adobe Fangsong Std}]{SimSun} 24 | \setCJKfamilyfont{zhsong}{SimSun} 25 | %\setCJKfamilyfont{zhhei}{Adobe Song Std} 26 | \setCJKfamilyfont{zhhei}{SimHei} 27 | %\setCJKfamilyfont{zhhei}{Adobe Heiti Std} 28 | \setCJKfamilyfont{zhkai}{KaiTi} 29 | \setCJKfamilyfont{zhfs}{FangSong} 30 | \usetheme{berkeley} 31 | 32 | \parindent 2em 33 | 34 | %title 35 | \title{生物智能与算法} 36 | \author{袁昕} 37 | \date{\today} 38 | 39 | \begin{document} 40 | 41 | \frame{\titlepage} 42 | 43 | %tables 44 | 45 | \section*{大纲} 46 | 47 | \frame{\tableofcontents} 48 | 49 | %summary 50 | 51 | \section{思考} 52 | 53 | \frame{\frametitle{思考} 54 | \begin{enumerate} 55 | \item<1-> 针对的问题是什么,除了优化问题和预测问题,还能处理其他什么问题? 56 | \item<2-> 如何对求解的量设计编码? 57 | \item<3-> 这一类方法的几何和物理本质是什么? 58 | \item<4-> 存在过度成熟问题吗?其优缺点和适用范围? 59 | \item<5-> 近年有哪些新的变种算法和模型? 60 | \item<6-> 有没有进一步改进的空间? 61 | \end{enumerate} 62 | } 63 | 64 | \frame{\frametitle{应用的最优化问题的形式} 65 | 应用的最优化问题的形式主要包括非凸函数优化和组合优化。 66 | 67 | ~ 68 | 69 | 也可用于复杂形式的凸函数优化,或者是在数值计算上呈病态的凸函数优化。 70 | } 71 | 72 | %topics 73 | 74 | \input{topics/evolution.tex} 75 | 76 | \input{topics/immunity.tex} 77 | 78 | \input{topics/swarm.tex} 79 | 80 | \input{topics/forage.tex} 81 | 82 | \input{topics/ann.tex} 83 | 84 | \input{topics/cn.tex} 85 | 86 | %end 87 | 88 | \frame{\frametitle{结束} 89 | \centerline{\textbf{\Huge{结束!}}} 90 | } 91 | 92 | \frame{\frametitle{版权申明} 93 | 94 | 本作品采用知识共享 署名-非商业性使用-禁止演绎 3.0 中国大陆 许可协议进行许可。 95 | 要查看该许可协议,可访问 http://creativecommons.org/licenses/by-nc-nd/3.0/cn/ 或者 96 | 写信到 Creative Commons, PO Box 1866, Mountain View, CA 94042, USA。 97 | } 98 | 99 | \end{document} 100 | -------------------------------------------------------------------------------- /slides/3.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \documentclass[11pt]{beamer} 6 | \usepackage{beamerthemesplit} 7 | \usepackage[BoldFont,SlantFont,CJKchecksingle]{xeCJK} 8 | \setCJKmainfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 9 | \setCJKsansfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 10 | \setCJKmonofont[ItalicFont={Adobe Fangsong Std}]{SimSun} 11 | \setCJKfamilyfont{zhsong}{SimSun} 12 | %\setCJKfamilyfont{zhhei}{Adobe Song Std} 13 | \setCJKfamilyfont{zhhei}{SimHei} 14 | %\setCJKfamilyfont{zhhei}{Adobe Heiti Std} 15 | \setCJKfamilyfont{zhkai}{KaiTi} 16 | \setCJKfamilyfont{zhfs}{FangSong} 17 | \usetheme{Madrid} 18 | \usecolortheme{crane} 19 | 20 | \parindent 2em 21 | 22 | %title 23 | \title{生物智能与算法} 24 | \author{袁昕} 25 | \date{\today} 26 | 27 | \begin{document} 28 | 29 | \frame{\titlepage} 30 | 31 | \frame{ 32 | \centerline{\textbf{\Huge{总结和展望}}} 33 | } 34 | 35 | \frame{\frametitle{能解决的问题} 36 | \begin{itemize} 37 | \item<1-> 确定性的优化问题、预测问题,准确率有限,漏报率和准确率之间有关系。 38 | \item<2-> 多数只能找到次优解。对次优解的评价没有统一的标准。 39 | \item<3-> 可能比较适合研究问题本身在不断变化的问题,即环境不断变化, 40 | 如视觉认知,不同条件下成像的猫。 41 | \end{itemize} 42 | } 43 | 44 | \frame{\frametitle{研究思路} 45 | \begin{itemize} 46 | \item<1-> 各个已有生物智能方法之间的交叉结合。 47 | \item<2-> 与描述复杂、不确定现象的各种理论结合。 48 | \item<3-> 吸收生物学研究、脑科学研究的最新成果。 49 | \end{itemize} 50 | } 51 | 52 | \frame{\frametitle{理论研究} 53 | \begin{itemize} 54 | \item<1-> 研究复杂网络非线性动力学模型、随机过程。 55 | \item<2-> 研究多智能主体间的作用关系,包括兴奋和抑制、合作和竞争、正负反馈。 56 | \item<3-> 网络内部的节点和连接代表的事物可变,其作用规律也可变,研究异质节点连接的网络。 57 | \item<4-> 研究网络的结构不断变化的情形(时序网络,自适应网络)。 58 | \item<5-> 研究网络的网络,高阶网络,多网络间的相互作用。 59 | \end{itemize} 60 | } 61 | 62 | \frame{\frametitle{应用研究} 63 | \begin{itemize} 64 | \item<1-> 快速并行算法。 65 | \item<2-> 快速收敛算法。 66 | \item<3-> 逻辑推理的应用。 67 | \item<4-> 视觉认知的应用。 68 | \item<5-> 听觉认知的应用。 69 | \item<6-> 智慧城市的应用。 70 | \item<7-> 互联网和安全的应用。 71 | \end{itemize} 72 | } 73 | 74 | \frame{\frametitle{应用基础研究} 75 | 应用基础研究即转化研究。 76 | 77 | ~ 78 | 79 | 要研究和开创工业上有用的理论和方法,需要将工程上可以测量和控制的变量,如温度,压力,变形等作为所求函数的自变量。 80 | } 81 | 82 | %end 83 | 84 | \frame{\frametitle{结束} 85 | \centerline{\textbf{\Huge{结束!}}} 86 | } 87 | 88 | \frame{\frametitle{版权申明} 89 | 90 | 本作品采用知识共享 署名-非商业性使用-禁止演绎 3.0 中国大陆 许可协议进行许可。 91 | 要查看该许可协议,可访问 http://creativecommons.org/licenses/by-nc-nd/3.0/cn/ 或者 92 | 写信到 Creative Commons, PO Box 1866, Mountain View, CA 94042, USA。 93 | } 94 | 95 | \end{document} 96 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | ## Core latex/pdflatex auxiliary files: 2 | *.aux 3 | *.lof 4 | *.log 5 | *.lot 6 | *.fls 7 | *.out 8 | *.toc 9 | *.fmt 10 | *.fot 11 | *.cb 12 | *.cb2 13 | 14 | ## Intermediate documents: 15 | *.dvi 16 | *-converted-to.* 17 | # these rules might exclude image files for figures etc. 18 | # *.ps 19 | # *.eps 20 | # *.pdf 21 | 22 | ## Bibliography auxiliary files (bibtex/biblatex/biber): 23 | *.bbl 24 | *.bcf 25 | *.blg 26 | *-blx.aux 27 | *-blx.bib 28 | *.brf 29 | *.run.xml 30 | 31 | ## Build tool auxiliary files: 32 | *.fdb_latexmk 33 | *.synctex 34 | *.synctex.gz 35 | *.synctex.gz(busy) 36 | *.pdfsync 37 | 38 | ## Auxiliary and intermediate files from other packages: 39 | # algorithms 40 | *.alg 41 | *.loa 42 | 43 | # achemso 44 | acs-*.bib 45 | 46 | # amsthm 47 | *.thm 48 | 49 | # beamer 50 | *.nav 51 | *.snm 52 | *.vrb 53 | 54 | # cprotect 55 | *.cpt 56 | 57 | # fixme 58 | *.lox 59 | 60 | #(r)(e)ledmac/(r)(e)ledpar 61 | *.end 62 | *.?end 63 | *.[1-9] 64 | *.[1-9][0-9] 65 | *.[1-9][0-9][0-9] 66 | *.[1-9]R 67 | *.[1-9][0-9]R 68 | *.[1-9][0-9][0-9]R 69 | *.eledsec[1-9] 70 | *.eledsec[1-9]R 71 | *.eledsec[1-9][0-9] 72 | *.eledsec[1-9][0-9]R 73 | *.eledsec[1-9][0-9][0-9] 74 | *.eledsec[1-9][0-9][0-9]R 75 | 76 | # glossaries 77 | *.acn 78 | *.acr 79 | *.glg 80 | *.glo 81 | *.gls 82 | *.glsdefs 83 | 84 | # gnuplottex 85 | *-gnuplottex-* 86 | 87 | # hyperref 88 | *.brf 89 | 90 | # knitr 91 | *-concordance.tex 92 | # TODO Comment the next line if you want to keep your tikz graphics files 93 | *.tikz 94 | *-tikzDictionary 95 | 96 | # listings 97 | *.lol 98 | 99 | # makeidx 100 | *.idx 101 | *.ilg 102 | *.ind 103 | *.ist 104 | 105 | # minitoc 106 | *.maf 107 | *.mlf 108 | *.mlt 109 | *.mtc 110 | *.mtc[0-9] 111 | *.mtc[1-9][0-9] 112 | 113 | # minted 114 | _minted* 115 | *.pyg 116 | 117 | # morewrites 118 | *.mw 119 | 120 | # mylatexformat 121 | *.fmt 122 | 123 | # nomencl 124 | *.nlo 125 | 126 | # sagetex 127 | *.sagetex.sage 128 | *.sagetex.py 129 | *.sagetex.scmd 130 | 131 | # sympy 132 | *.sout 133 | *.sympy 134 | sympy-plots-for-*.tex/ 135 | 136 | # pdfcomment 137 | *.upa 138 | *.upb 139 | 140 | # pythontex 141 | *.pytxcode 142 | pythontex-files-*/ 143 | 144 | # thmtools 145 | *.loe 146 | 147 | # TikZ & PGF 148 | *.dpth 149 | *.md5 150 | *.auxlock 151 | 152 | # todonotes 153 | *.tdo 154 | 155 | # xindy 156 | *.xdy 157 | 158 | # xypic precompiled matrices 159 | *.xyc 160 | 161 | # endfloat 162 | *.ttt 163 | *.fff 164 | 165 | # Latexian 166 | TSWLatexianTemp* 167 | 168 | ## Editors: 169 | # WinEdt 170 | *.bak 171 | *.sav 172 | 173 | # Texpad 174 | .texpadtmp 175 | 176 | # Kile 177 | *.backup 178 | 179 | # KBibTeX 180 | *~[0-9]* 181 | -------------------------------------------------------------------------------- /slides/2/topics/ann.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{人工神经网络} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{人工神经网络}}} 9 | } 10 | 11 | \frame{\frametitle{定义} 12 | 13 | 人工神经网络是由大量处理单元互联组成的非线性、自适应信息处理系统。 14 | } 15 | 16 | \frame{\frametitle{思想} 17 | 18 | 在现代神经科学研究成果的基础上提出,试图通过模拟大脑神经网络处理、记忆信息的方式进行信息处理。 19 | 20 | ~ 21 | 22 | 非线性、非局限性、非定常性、非凸性。 23 | 24 | ~ 25 | 26 | 自适应、自组织、自学习。 27 | } 28 | 29 | \frame{\frametitle{思想} 30 | 人工智能的三大流派: 31 | 32 | \begin{itemize} 33 | \item 符号主义 34 | \item 联结主义 35 | \item 行为主义 36 | \end{itemize} 37 | } 38 | 39 | \frame{\frametitle{思想} 40 | 41 | 基于逻辑符号和规则推理的专家系统,在处理直觉、非结构化信息方面有缺陷。 42 | 43 | ~ 44 | 45 | 人工神经网络是连接主义的观点,是神经元相互联接而成的自适应非线性动态系统, 46 | 分为前向网络(有向无环图)和反馈网络(无向完备图)两类。 47 | 48 | ~ 49 | 50 | 理论研究:MP模型、Hebb规则、感知器、适应谐振ART理论、自组织映射、认知机网络、非线性动力学。 51 | } 52 | 53 | \frame{\frametitle{思想} 54 | \begin{itemize} 55 | \item<1-> 万能逼近理论、高维空间中的低维流形、类比分段函数、 56 | 映射到高维特征空间(重新部署点以便获得线性处理或其他性质良好的处理)、 57 | 高阶到低阶的局部调制。 58 | \item<2-> 泛化问题:欠拟合,过拟合,正则的作用,VC维模型复杂度理论, 59 | 训练集和验证集的复杂度-误差-方差曲线 60 | \item<3-> 凸优化与非凸优化:加凸集约束的正则项(模型复杂程度的单调增函数,几何解释); 61 | 梯度消失,梯度爆炸,自动微分 62 | \item<4-> 对抗网络的本质,边界的确定,与正则的关系,与传输方程的关系。 63 | \end{itemize} 64 | } 65 | 66 | \frame{\frametitle{种类} 67 | \begin{itemize} 68 | \item<1-> Hopfield 69 | \item<2-> 波尔兹曼机(模拟退火及MCMC) 70 | \item<3-> 联想记忆 71 | \item<4-> BP 72 | \item<5-> 极限学习机(ELM) 73 | \item<6-> 局部逼近神经网络(CMAC,RBF,B样条) 74 | \end{itemize} 75 | } 76 | 77 | \frame{\frametitle{种类} 78 | \begin{itemize} 79 | \item<1-> 自组织SOM 80 | \item<2-> ART 81 | \item<3-> SVM(描述模型复杂度的VC维,核方法,升维) 82 | \item<4-> 强化学习(行为主义,可分层,多智能体,可类比自动控制) 83 | \item<5-> 降维处理(主成分分析,低秩,补全问题) 84 | \end{itemize} 85 | } 86 | 87 | \frame{\frametitle{种类} 88 | \begin{itemize} 89 | \item<1-> 集成式网络(boosting,梯度推进机,决策树,随机森林,知识图谱) 90 | \item<2-> 量子神经网络 91 | \item<3-> 脉冲耦合 92 | \item<4-> 混沌神经网络 93 | \item<5-> 深度学习(卷积、循环、递归,置信网络,生成式对抗网络,编解码器,迁移学习,教师学生,蒸馏, 94 | 端到端,注意力,预训练,扩散模型,对比学习,增量学习,果蝇的大脑) 95 | \end{itemize} 96 | } 97 | 98 | \frame{\frametitle{种类} 99 | \begin{itemize} 100 | \item<1-> 图神经网络 101 | \item<2-> 隐私保护学习(联邦学习,差分隐私学习,联合学习,分离学习) 102 | \item<3-> 因果学习(关联、干预、反事实估计和推断,结构因果,因果强化,去相关,深度稳定学习、方差分析) 103 | \item<4-> 大语言模型 104 | \end{itemize} 105 | } 106 | 107 | \frame{\frametitle{深度学习} 108 | 深度学习的科普 109 | 110 | ~ 111 | 112 | 参见\href{https://ishare.ifeng.com/c/s/v002JgwF4NYcbmQ82S4-_eS-_3--f5R3EL7B4JxV5Rtmq7LCNA__}{网页} 113 | } 114 | 115 | \frame{\frametitle{随机优化} 116 | 针对的是使用梯度下降法求解最优化目标泛函的问题。 117 | 其中的训练样本数据都是最优化目标泛函中的常数。 118 | 具体研究的是训练样本数量过大的情形下梯度的计算问题。 119 | 120 | ~ 121 | 122 | 从一个初始的点(代表一个函数的参数)出发在函数空间中搜索最优化目标函数的解。 123 | 每次迭代都只选用部分样本数据或者重新采样来估计当前的各种梯度来更新解, 124 | 以减少计算量。 125 | } 126 | 127 | \frame{\frametitle{随机优化} 128 | \begin{itemize} 129 | \item<1-> 零阶估计(用函数值和一些假设算梯度和二阶海森矩阵) 130 | \item<2-> 方差缩减(每次样本采样,梯度值不一样,要看成随机变量,方差大,结合先前迭代步信息缩减) 131 | \item<3-> 二阶稳定点(二阶海森矩阵,抑制收敛到局部最大点,局部鞍点) 132 | \item<4-> 动量(每次再多走一点距离,加快收敛) 133 | \item<5-> 朗之万动力学(随机噪声) 134 | \end{itemize} 135 | } 136 | 137 | \frame{\frametitle{随机优化} 138 | \begin{itemize} 139 | \item<1-> 假设依然较多,为了证明收敛,更多是在凑数字的游戏。适用场景存疑。 140 | \item<2-> 每次迭代的对当前点的修改量,或者求下一个点,都要求解一个当前点周边范围内的最优化子问题。 141 | \item<3-> 依然无法避免收敛到局部最小点。 142 | \end{itemize} 143 | } 144 | 145 | \frame{\frametitle{阅读} 146 | \begin{itemize} 147 | \item<1-> 反思。参见\href{https://mp.weixin.qq.com/s/hdBGdOGcwxDt8mfCjj4mQQ}{网页} 148 | \item<2-> 现状。参见\href{https://mp.weixin.qq.com/s/2YC1N3MB8G7mw1HThucceQ}{网页} 149 | \item<3-> 工业案例。参见\href{https://mp.weixin.qq.com/s/YTI5QxNjVIq2PJeOKxbc8g}{网页} 150 | \item<4-> 其他落地案例:非关键、消费级,如电子商务推荐等。 151 | \end{itemize} 152 | } 153 | 154 | %end 155 | -------------------------------------------------------------------------------- /slides/1.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2018 3 | % 4 | 5 | \documentclass[11pt]{beamer} 6 | \usepackage{beamerthemesplit} 7 | %url 8 | \usepackage{hyperref} 9 | \hypersetup{colorlinks=true, urlcolor=blue} 10 | %xeCJK 11 | \usepackage[BoldFont,SlantFont,CJKchecksingle]{xeCJK} 12 | \setCJKmainfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 13 | \setCJKsansfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 14 | \setCJKmonofont[ItalicFont={Adobe Fangsong Std}]{SimSun} 15 | \setCJKfamilyfont{zhsong}{SimSun} 16 | %\setCJKfamilyfont{zhhei}{Adobe Song Std} 17 | \setCJKfamilyfont{zhhei}{SimHei} 18 | %\setCJKfamilyfont{zhhei}{Adobe Heiti Std} 19 | \setCJKfamilyfont{zhkai}{KaiTi} 20 | \setCJKfamilyfont{zhfs}{FangSong} 21 | \usetheme{goettingen} 22 | \usecolortheme{beetle} 23 | 24 | \parindent 2em 25 | 26 | %title 27 | \title{生物智能与算法} 28 | \author{袁昕} 29 | \date{\today} 30 | 31 | \begin{document} 32 | 33 | \frame{\titlepage} 34 | 35 | \frame{ 36 | \centerline{\textbf{\Huge{基础知识}}} 37 | } 38 | 39 | %tables 40 | 41 | \section*{大纲} 42 | 43 | \frame{\tableofcontents} 44 | 45 | %summary 46 | 47 | \section{基础} 48 | 49 | \frame{\frametitle{科普} 50 | \begin{itemize} 51 | \item<1-> 科学网,果壳,科普中国等 52 | \item<2-> 图形和视频摘要 53 | \item<3-> 折纸 54 | \end{itemize} 55 | } 56 | 57 | \frame{\frametitle{计算智能的早期发展} 58 | \begin{itemize} 59 | \item<1-> 自动控制的故事 60 | \item<2-> 系统辨识(带有激励的数据实验) 61 | \item<3-> 工业大数据是否有效的辨析 62 | \item<4-> 一般智能:从模式识别到人工智能 63 | \end{itemize} 64 | } 65 | 66 | \frame{\frametitle{自动控制的知识体系} 67 | \begin{itemize} 68 | \item<1-> 自动化学科:参见\href{https://blog.sciencenet.cn/blog-3387802-1331861.html}{网页} 69 | \item<2-> 图示:参见\href{https://blog.sciencenet.cn/blog-3387802-1331871.html}{网页} 70 | \item<3-> 基本问题:参见\href{https://blog.sciencenet.cn/blog-3387802-1331872.html}{网页} 71 | \item<4-> 多智能体和复杂网络:参见\href{https://blog.sciencenet.cn/blog-3387802-1380777.html}{网页} 72 | \item<5-> 理论和工程:参见\href{https://blog.sciencenet.cn/blog-3316223-1368152.html}{网页} 73 | \end{itemize} 74 | } 75 | 76 | %optimization 77 | 78 | \section{最优化理论} 79 | 80 | \frame{\frametitle{最优化} 81 | \begin{itemize} 82 | \item<1-> 问题描述 83 | \item<2-> 求解方法 84 | \item<3-> 正则化的本质(约束是什么,假设有噪声引起偏离) 85 | \item<4-> 泛函角度 86 | \item<5-> 物理角度 87 | \item<6-> 几何直观角度(微分几何) 88 | \item<7-> 统计角度(频率、贝叶斯、公理) 89 | \end{itemize} 90 | } 91 | 92 | \frame{\frametitle{问题描述} 93 | \begin{itemize} 94 | \item<1->凸函数和非凸函数 95 | \item<2->线性函数和非线性函数 96 | \item<3->有限维空间和无限维空间 97 | \item<4->离散函数(如组合优化)和连续函数(能量泛函及最大似然) 98 | \end{itemize} 99 | } 100 | 101 | \frame{\frametitle{拉格朗日乘子法} 102 | 几何直观解释: 103 | 104 | ~ 105 | 106 | 参见\href{http://blog.sciencenet.cn/blog-3428464-1283235.html}{网页} 107 | } 108 | 109 | \frame{\frametitle{正则化} 110 | 几何直观解释: 111 | 112 | ~ 113 | 114 | 参见\href{https://mp.weixin.qq.com/s?__biz=MzU0MDQ1NjAzNg==&mid=2247541375&idx=1&sn=2c4e882d69e0cbec156eec71a38cf599&chksm=fb3a8774cc4d0e62899bb78208be93d1bd2fe3798090c00c9a9ae078a2cd3bd98a6444594cb4&scene=27}{网页} 115 | } 116 | 117 | \frame{\frametitle{求解} 118 | \begin{itemize} 119 | \item<1-> 传统的数值优化算法。 120 | \item<2-> 现代的生物智能优化算法。 121 | \end{itemize} 122 | } 123 | 124 | \frame{\frametitle{求解-传统} 125 | 传统的数值优化算法,以严格的数学原理为指导,通过多次迭代,不断逼近全局最优解的算法, 126 | 比如基于梯度的算法、几何规划算法和二次规划算法等等。 127 | 这类算法的求解过程可以简述如下:首先,依据经验选定一个初始方案, 128 | 然后依据严格的数学原理来对初始方案进行不断的迭代,达到最大迭代次数,算法就停止了, 129 | 所获取的最终方案,即为找到的可能的全局最优方案。显然,这类算法对于初始方案的依赖性非常强。 130 | 鉴于数学原理推导的缜密性,一旦初始方案选定之后,最终方案实际上已经确定了。 131 | 因此,如果初始方案选择的较为合适,会得到一个比较理想的解决方案;否则,则会得到一个糟糕的解决方案。 132 | 此外,这类算法通常还会需要梯度信息,会大大增加求解的复杂程度。 133 | } 134 | 135 | \frame{\frametitle{求解-现代} 136 | 现代的生物智能优化算法,是通过随机性和简单规则的结合,通过模拟自然现象,来实现对优化问题的求解, 137 | 比如经典的粒子群优化算法和差分进化算法等等。 138 | 这类算法的求解过程可以简述如下:首先,对种群随机初始化,然后依据设定的简单规则来引导种群的搜索方向, 139 | 达到最大迭代次数了,算法就停止了,所获取的最终方案即为找到的可能的全局最优方案。 140 | 显然,这类算法由于在搜索机制中引入了随机数以及种群的随机初始化,就大大弱化了这类算法对于初始方案依赖程度。 141 | 此外,这类算法不需要梯度信息。 142 | } 143 | 144 | \frame{\frametitle{求解-对比} 145 | 在求解复杂优化问题的时候,相较于传统的数值优化算法,生物智能优化算法展现出更强的避免局部最优解的能力。 146 | 当前,生物智能优化算法已经作为一种非常主流的优化技术广泛应用不同领域的优化问题。 147 | 这一结论目前已经被发表的大量论文所证实。 148 | } 149 | 150 | %end 151 | 152 | \frame{\frametitle{结束} 153 | \centerline{\textbf{\Huge{结束!}}} 154 | } 155 | 156 | \frame{\frametitle{版权申明} 157 | 158 | 本作品采用知识共享 署名-非商业性使用-禁止演绎 3.0 中国大陆 许可协议进行许可。 159 | 要查看该许可协议,可访问 http://creativecommons.org/licenses/by-nc-nd/3.0/cn/ 或者 160 | 写信到 Creative Commons, PO Box 1866, Mountain View, CA 94042, USA。 161 | } 162 | 163 | \end{document} 164 | -------------------------------------------------------------------------------- /slides/2/topics/swarm.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \section{群聚智能} 6 | 7 | \frame{ 8 | \centerline{\textbf{\Huge{群聚智能}}} 9 | } 10 | 11 | \frame{\frametitle{定义} 12 | 13 | 无智能或简单智能的主体通过任何形式的聚集协作而表现出智能行为的特性。 14 | 15 | ~ 16 | 17 | 群,是一组相互之间可以进行直接或间接通信(通过改变局部环境)的主体。 18 | } 19 | 20 | \frame{\frametitle{思想} 21 | 22 | 源于分子自动机系统的自组织研究。 23 | 在没有集中且不提供全局模型的前提下,为寻找复杂分布式问题解决方案提供基础。 24 | } 25 | 26 | \frame{\frametitle{思想} 27 | 相对于传统的梯度优化算法 28 | 29 | ~ 30 | 31 | \begin{itemize} 32 | \item<1-> 完全分布式处理个体间和个体环境间作用,具有自组织性; 33 | \item<2-> 个体间非直接,合作基于环境感知,系统可扩展和安全; 34 | \item<3-> 无集中控制,具有鲁棒性,个别个体失效不会影响整个问题求解; 35 | \item<4-> 个体智能简单,实现方便,执行时间短。 36 | \end{itemize} 37 | } 38 | 39 | \frame{\frametitle{种类} 40 | \begin{itemize} 41 | \item<1-> 蚁群(蚁狮) 42 | \item<2-> 鸟群(粒子群) 43 | \item<3-> 人工鱼群 44 | \item<4-> 蜂群 45 | \item<5-> 蛙群 46 | \end{itemize} 47 | } 48 | 49 | \frame{\frametitle{种类} 50 | \begin{itemize} 51 | \item<1-> 萤火虫 52 | \item<2-> 蝙蝠群 53 | \item<3-> 布谷鸟 54 | \item<4-> 麻雀 55 | \item<5-> 社会性蜘蛛 56 | \end{itemize} 57 | } 58 | 59 | \frame{\frametitle{种类} 60 | \begin{itemize} 61 | \item<1-> 灰狼 62 | \item<2-> 猫群 63 | \item<3-> 狮群 64 | \item<4-> 斑点鬣狗 65 | \item<5-> 鲸鱼捕食 66 | \item<6-> 帝企鹅 67 | \item<7-> 黑猩猩 68 | \end{itemize} 69 | } 70 | 71 | \frame{\frametitle{种类} 72 | \begin{itemize} 73 | \item<1-> 磷虾 74 | \item<2-> 蟑螂 75 | \item<3-> 蝗虫 76 | \item<4-> 蝴蝶 77 | \item<5-> 飞蛾扑火 78 | \item<6-> 大红斑蝶 79 | \item<7-> 天牛须 80 | \end{itemize} 81 | } 82 | 83 | \frame{\frametitle{种类} 84 | \begin{itemize} 85 | \item<1-> 黏菌 86 | \item<2-> 蚯蚓 87 | \item<3-> 大象放牧 88 | \item<4-> 樽海鞘群 89 | \item<5-> 乌燕鸥 90 | \item<6-> 共生生物搜索 91 | \end{itemize} 92 | } 93 | 94 | \frame{\frametitle{种类} 95 | \begin{itemize} 96 | \item<1-> 花朵授粉 97 | \item<2-> 烟花 98 | \end{itemize} 99 | } 100 | 101 | \frame{\frametitle{公式示例} 102 | 103 | \begin{equation*} 104 | p_{ij}^{k}(t) = \left\{ 105 | \begin{array}{lc} 106 | \frac{[\tau_{ij}(t)]^{\alpha}\cdot[\eta_{ij}(t)]^{\beta}}{\sum\limits_{s\in J_{k}(i)} [\tau_{is}(t)]^{\alpha}\cdot[\eta_{is}(t)]^{\beta}}, & \mbox{如果} J \in J_{k}(i) \\ 107 | 0, & \mbox{否则} 108 | \end{array} \right. 109 | \end{equation*} 110 | } 111 | 112 | \frame{\frametitle{定理示例} 113 | 114 | \newtheorem{zh-thm}{定理} 115 | \begin{zh-thm} 116 | 如果有a,b,c, 则有$a^2+b^2=c^2$。 117 | \end{zh-thm} 118 | 119 | \renewcommand\proofname{证明} 120 | \begin{proof} 121 | 因为有a,b,c, 所以有$a^2+b^2=c^2$。 122 | \end{proof} 123 | } 124 | 125 | \frame{\frametitle{算法示例} 126 | 127 | \begin{algorithm}[H] 128 | \caption{计算 $y=x^n$}\label{alg1} 129 | \algsetup{linenosize=\tiny} \scriptsize 130 | \begin{algorithmic} 131 | \REQUIRE $n \geq 0 \vee x \neq 0$ 132 | \ENSURE $y=x^n$ 133 | \STATE $y \Leftarrow 1$ 134 | \IF{$n < 0$} 135 | \STATE $X \Leftarrow 1 / x$ 136 | \STATE $N \Leftarrow -n$ 137 | \ELSE 138 | \STATE $X \Leftarrow x$ 139 | \STATE $N \Leftarrow n$ 140 | \ENDIF 141 | \WHILE{$N \neq 0$} 142 | \IF{$N$ is even} 143 | \STATE $X \Leftarrow X \times X$ 144 | \STATE $N \Leftarrow N / 2$ 145 | \ELSE[$N$ is odd] 146 | \STATE $y \Leftarrow y \times X$ 147 | \STATE $N \Leftarrow N - 1$ 148 | \ENDIF 149 | \ENDWHILE 150 | \end{algorithmic} 151 | \end{algorithm} 152 | } 153 | 154 | \frame{\frametitle{流程图示例} 155 | 156 | \tikzstyle{block}=[rectangle,draw,fill=blue!20,text width=4em,text centered,rounded corners] 157 | \tikzstyle{huge-block}=[rectangle,draw,fill=cyan!20,text width=5em,text centered,rounded corners,minimum height=4em] 158 | \tikzstyle{line}=[draw] 159 | \begin{tikzpicture}[node distance=2cm, auto] 160 | \path[use as bounding box] (-4, 0) rectangle (10, -2); 161 | \path[line]<1-> node[huge-block](center){群聚智能}; 162 | \path[line, <->]<2-> node[block, below of=center, node distance=3cm](center-develop){蚁群算法} 163 | (center)--(center-develop); 164 | \path[line, <->]<3-> node[block, left of=center-develop, node distance=3cm](left-develop){粒子群} 165 | (center)--(left-develop); 166 | \path[line, <->]<4-> node[block, right of=center-develop, node distance=3cm](right-develop){灰狼} 167 | (center)--(right-develop); 168 | \end{tikzpicture} 169 | } 170 | 171 | \frame{\frametitle{流程图示例} 172 | 173 | \begin{tikzpicture}[font={\sf \scriptsize}] 174 | 175 | \node (start) at (-2,3) [draw, terminal, minimum width=2cm, minimum height=0.5cm] {开始}; 176 | \node (getdata) at (-2,2) [draw, predproc, align=left, minimum width=2cm, minimum height=0.5cm] {读取数据}; 177 | \node (init) at (-2,1) [draw, process, minimum width=3cm, minimum height=0.5cm] {设置参数和初始化}; 178 | \coordinate (point1) at (-0.25,1); 179 | \node (eval) at (1,1) [draw, process, minimum width=2cm, minimum height=0.5cm] {评价蚁群}; 180 | \node (decide) at (1,-1) [draw, decision, minimum width=2cm, minimum height=0.5cm] {判断终止条件}; 181 | \node (storage) at (1,-3) [draw, storage, minimum width=3cm, minimum height=0.5cm] {输出结果并存储}; 182 | \node (iter) at (4,0) [draw, process, minimum width=3cm, minimum height=0.5cm] {迭代次数增加}; 183 | \node (select) at (4,0.7) [draw, process, minimum width=3cm, minimum height=0.5cm] {概率选择移动方向}; 184 | \node (update) at (4,1.4) [draw, process, minimum width=3cm, minimum height=0.5cm] {信息素更新}; 185 | \coordinate (point2) at (4,2); 186 | \node (end) at (4,-3) [draw, terminal, minimum width=2cm, minimum height=0.5cm] {结束}; 187 | 188 | \draw[->] (start) -- (getdata); 189 | \draw[->] (getdata) -- (init); 190 | \draw[->] (init) -- (point1) -- (eval); 191 | \draw[->] (eval) -- (decide); 192 | \draw[->] (decide) -| node[right]{否} (iter); 193 | \draw[->] (decide) -- node[right]{是} (storage); 194 | \draw[->] (iter) -- (select); 195 | \draw[->] (select) -- (update); 196 | \draw[->] (update) -- (point2) -| (point1); 197 | \draw[->] (storage) -- (end); 198 | \end{tikzpicture} 199 | } 200 | 201 | \frame{\frametitle{动画示例} 202 | \begin{figure}[hb] 203 | \centering 204 | \animategraphics[height=2.5in, autoplay, controls]{1}{topics/swarm-files/lzq_}{0}{5} 205 | \end{figure} 206 | } 207 | 208 | %end 209 | -------------------------------------------------------------------------------- /slides/4.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2023 3 | % 4 | 5 | \documentclass[11pt]{beamer} 6 | \usepackage{beamerthemesplit} 7 | %url 8 | \usepackage{hyperref} 9 | \hypersetup{colorlinks=true, urlcolor=blue} 10 | %xeCJK 11 | \usepackage[BoldFont,SlantFont,CJKchecksingle]{xeCJK} 12 | \setCJKmainfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 13 | \setCJKsansfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 14 | \setCJKmonofont[ItalicFont={Adobe Fangsong Std}]{SimSun} 15 | \setCJKfamilyfont{zhsong}{SimSun} 16 | %\setCJKfamilyfont{zhhei}{Adobe Song Std} 17 | \setCJKfamilyfont{zhhei}{SimHei} 18 | %\setCJKfamilyfont{zhhei}{Adobe Heiti Std} 19 | \setCJKfamilyfont{zhkai}{KaiTi} 20 | \setCJKfamilyfont{zhfs}{FangSong} 21 | \usetheme{goettingen} 22 | \usecolortheme{beetle} 23 | 24 | \parindent 2em 25 | 26 | %title 27 | \title{生物智能与算法} 28 | \author{袁昕} 29 | \date{\today} 30 | 31 | \begin{document} 32 | 33 | \frame{\titlepage} 34 | 35 | \frame{ 36 | \centerline{\textbf{\Huge{理论提升}}} 37 | } 38 | 39 | %tables 40 | 41 | \section*{大纲} 42 | 43 | \frame{\tableofcontents} 44 | 45 | %physics 46 | 47 | \section{数学与物理} 48 | 49 | \frame{\frametitle{数学和物理} 50 | 物理理论和数学描述之间的关系: 51 | 参见\href{http://blog.sciencenet.cn/blog-39419-1271041.html}{网页} 52 | 53 | ~ 54 | 55 | 全局尺度与局部尺度: 56 | 参见\href{http://blog.sciencenet.cn/blog-39419-1273157.html}{网页} 57 | } 58 | 59 | \frame{\frametitle{物理的历史发展和思维} 60 | \begin{itemize} 61 | \item<1-> 从牛顿力学到相对论:参见\href{https://mp.weixin.qq.com/s/amhuQYFELB8gmhr2d5XgMg}{网页} 62 | \item<2-> 深刻理解侠义相对论:参见\href{https://blog.sciencenet.cn/blog-3377-1325424.html}{网页} 63 | \item<3-> 郎道理解狭义相对论:参见\href{https://blog.sciencenet.cn/blog-3377-1326910.html}{网页} 64 | \item<4-> 能量守恒:参见\href{http://blog.sciencenet.cn/blog-39346-1293732.html}{网页} 65 | \item<5-> 经典力学:参见\href{https://blog.sciencenet.cn/blog-709494-1353458.html}{网页} 66 | \end{itemize} 67 | } 68 | 69 | %books 70 | 71 | \section{数学教材} 72 | 73 | \frame{\frametitle{教科书} 74 | \begin{itemize} 75 | \item<1-> 微积分、集合到拓扑 76 | \item<2-> 集合、测度、概率论、统计到实分析 77 | \item<3-> 平面几何、解析几何、线性代数到向量代数 78 | \item<4-> 初等代数、复杂微积分、一维到多维面积计算、复数 79 | \item<5-> 复数、向量空间、超越空间、射影几何、无穷到有限、黎曼球面到复分析 80 | \item<6-> 有限和无限、傅里叶级数、泛函分析、傅里叶变换、对偶空间、组合数学到最优化理论 81 | \end{itemize} 82 | } 83 | 84 | \frame{\frametitle{教科书} 85 | \begin{itemize} 86 | \item<1-> 函数映射的一对多和多对一,空间的扭曲回环,非线性,空间嵌入,到升维和降维处理。 87 | \item<2-> 高阶函数、张量代数,流形,到泛函。 88 | \item<3-> 随机过程、随机微分方程、组合数学、离散数学、运筹学到博弈论。 89 | \item<4-> 数理金融、数学工具与实验观察不符、质疑和倒查、用平稳研究的非平稳,随机变量和样本轨道可能搞混。 90 | \item<5-> 抽象代数、微分几何到代数几何。 91 | \end{itemize} 92 | } 93 | 94 | %mathematics 95 | 96 | \section{提升数学水平} 97 | 98 | \frame{\frametitle{数学体系} 99 | \begin{enumerate} 100 | \item<1-> 数学体系,参见\href{https://mp.weixin.qq.com/s/Jo7ID5xl8eOOFvUcGh8Umw}{网页} 101 | \item<2-> 数理逻辑两种方法要结合,参见\href{https://blog.sciencenet.cn/blog-692836-1373598.html}{网页} 102 | \end{enumerate} 103 | } 104 | 105 | \frame{\frametitle{如何学好高数} 106 | 如何学好高数 107 | 108 | ~ 109 | 110 | 参见\href{https://mp.weixin.qq.com/s/a2wpf7sj80IXSQwEbs_S9A}{网页} 111 | } 112 | 113 | \frame{\frametitle{数学即数数或数圈} 114 | \begin{enumerate} 115 | \item<1-> 数得准即代数 116 | \item<2-> 数不准即分析(几何辅助) 117 | \item<3-> 数不清即几何拓扑(无穷小,无穷大,说不清,概率统计) 118 | \item<4-> 怎么数即逻辑 119 | \end{enumerate} 120 | } 121 | 122 | \frame{\frametitle{关于线性代数} 123 | 线性代数的矩阵(对比解析几何;处理坐标系无关的量) 124 | 125 | ~ 126 | 127 | 参见\href{https://mp.weixin.qq.com/s/DB6_w-TuaFCxR6Fws8l99w}{网页} 128 | } 129 | 130 | \frame{\frametitle{关于统计} 131 | \begin{enumerate} 132 | \item<1-> 统计学思想总结,参见\href{https://mp.weixin.qq.com/s/Shbhd2SaAg2DmikrDdOuSw}{网页} 133 | \item<2-> 概率的意义:随机世界与大数法则,参见\href{https://mp.weixin.qq.com/s/yxyaEcMolGMen3-65nWCdg}{网页} 134 | \item<3-> 测量误差与统计推断问题的统一性,参见\href{https://blog.sciencenet.cn/blog-103568-1346732.html}{网页} 135 | \end{enumerate} 136 | } 137 | 138 | \frame{\frametitle{关于几何} 139 | 微分几何 140 | 141 | ~ 142 | 143 | 参见\href{https://mp.weixin.qq.com/s/nk1ta8jVrqubwhSHekq0tA}{网页} 144 | } 145 | 146 | \frame{\frametitle{不确定性} 147 | 为何要融合Mixture?参见\href{http://blog.sciencenet.cn/blog-388372-1299920.html}{网页} 148 | 149 | ~ 150 | 151 | 两大载体:变量、模型。 152 | } 153 | 154 | \frame{\frametitle{不确定性} 155 | \begin{enumerate} 156 | \item<1-> 现实世界不可知变量的不确定性往往用noise建模,如系统噪声、量测噪声等。 157 | \item<2-> 不确定的模型或者假设往往混合mixture形式来表达,如高斯混合GMM,多模型MM,交互多模型IMM, 158 | 多假设跟踪器MHT,蒙特卡洛方法(粒子群混合),随机有限集统计学框架下的多伯努利混合MBM,广义标签多伯努利GLMB,等。 159 | 还有Watson mixture,inverted Beta mixture,von-Mises Fisher mixture model, 160 | 以及复合型混合:Gaussian-Student's-t mixture,Gaussian-uniform mixture等。 161 | \end{enumerate} 162 | } 163 | 164 | \frame{\frametitle{不确定性} 165 | 思想是多模型/多分布加权平均(Arithmetic average),即AA融合模型, 166 | 也与神经网络neuron network(NN)有着内在关联(NN可看作大量神经元的mixture混合工作机制)。 167 | 可以进一步从几何角度思考。 168 | } 169 | 170 | \frame{\frametitle{应用数学} 171 | 应用数学的强大威力 172 | 173 | ~ 174 | 175 | 参见\href{https://mp.weixin.qq.com/s/ks05jYJHG-bleVcPhDWQhA}{网页} 176 | } 177 | 178 | \frame{\frametitle{探讨} 179 | \begin{enumerate} 180 | \item<1-> 微分和导数,是线性化的开端,结合了无穷的思想, 181 | 参见\href{http://blog.sciencenet.cn/blog-3377-1289944.html}{网页} 182 | \item<2-> 概率定义的探讨,相对频率、个人主观信念度、倾向度。 183 | 参见\href{https://blog.sciencenet.cn/blog-3503579-1321302.html}{网页} 184 | \item<3-> 随机过程,参见\href{http://blog.sciencenet.cn/home.php?mod=space&uid=3418723&do=blog&id=1286356}{网页1} 185 | \href{https://blog.sciencenet.cn/blog-3418723-1365894.html}{网页2} 186 | \end{enumerate} 187 | } 188 | 189 | %philosophy 190 | 191 | \section{科学哲学} 192 | 193 | \frame{\frametitle{经典科学哲学思路} 194 | \begin{itemize} 195 | \item<1-> 还原论 196 | \item<2-> 因果论(穆勒五法) 197 | \item<3-> 逻辑学(同一律,排中律,矛盾律,因果律,区分对立和矛盾的概念) 198 | \item<4-> 本体论 199 | \item<5-> 科学研究的范式(数据驱动,开普勒范式;基本原理驱动,牛顿范式) 200 | \end{itemize} 201 | } 202 | 203 | \frame{\frametitle{现代科学哲学思路} 204 | \begin{itemize} 205 | \item<1-> 整体论 206 | \item<2-> 系统论(结合控制论,讨论所有可能的变量和规律,调整事物之间的关系) 207 | \item<3-> 层次论(从高阶函数、泛函升级) 208 | \item<4-> 相关性和因果性(数据驱动和原理驱动,因果是幻觉吗?) 209 | \item<5-> 互为因果和循环论证(宗教的神逻辑) 210 | \end{itemize} 211 | } 212 | 213 | \frame{\frametitle{科学哲学} 214 | 一流的科学家对哲学也应有深刻的认识,参见\href{http://blog.sciencenet.cn/blog-3387120-1273402.html}{网页} 215 | 216 | ~ 217 | 218 | 库恩科学哲学范式理论,参见\href{http://blog.sciencenet.cn/blog-71079-1273375.html}{网页} 219 | 220 | ~ 221 | 222 | 波普尔的理论,参见\href{http://blog.sciencenet.cn/blog-1255140-769490.html}{网页} 223 | } 224 | 225 | \frame{\frametitle{科学难题} 226 | 科学难题的形成 227 | 228 | ~ 229 | 230 | 参见\href{https://blog.sciencenet.cn/blog-39419-1272556.html}{网页} 231 | } 232 | 233 | %end 234 | 235 | \frame{\frametitle{结束} 236 | \centerline{\textbf{\Huge{结束!}}} 237 | } 238 | 239 | \frame{\frametitle{版权申明} 240 | 241 | 本作品采用知识共享 署名-非商业性使用-禁止演绎 3.0 中国大陆 许可协议进行许可。 242 | 要查看该许可协议,可访问 http://creativecommons.org/licenses/by-nc-nd/3.0/cn/ 或者 243 | 写信到 Creative Commons, PO Box 1866, Mountain View, CA 94042, USA。 244 | } 245 | 246 | \end{document} 247 | -------------------------------------------------------------------------------- /slides/0.tex: -------------------------------------------------------------------------------- 1 | % 2 | % GNU courseware, XIN YUAN, 2017 3 | % 4 | 5 | \documentclass[11pt]{beamer} 6 | \usepackage{beamerthemesplit} 7 | \usepackage{graphicx} 8 | \usepackage{colortbl} 9 | \usepackage[BoldFont,SlantFont,CJKchecksingle]{xeCJK} 10 | \setCJKmainfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 11 | \setCJKsansfont[BoldFont=SimHei,SlantedFont=KaiTi]{SimSun} 12 | \setCJKmonofont[ItalicFont={Adobe Fangsong Std}]{SimSun} 13 | \setCJKfamilyfont{zhsong}{SimSun} 14 | %\setCJKfamilyfont{zhhei}{Adobe Song Std} 15 | \setCJKfamilyfont{zhhei}{SimHei} 16 | %\setCJKfamilyfont{zhhei}{Adobe Heiti Std} 17 | \setCJKfamilyfont{zhkai}{KaiTi} 18 | \setCJKfamilyfont{zhfs}{FangSong} 19 | \setCJKfamilyfont{zhls}{LiSu} 20 | \setCJKfamilyfont{zhyy}{YouYuan} 21 | \usetheme{warsaw} 22 | 23 | \parindent 2em 24 | 25 | %title 26 | \title{生物智能与算法} 27 | \author{袁昕} 28 | \date{\today} 29 | 30 | \begin{document} 31 | 32 | %title 33 | \frame{\titlepage} 34 | 35 | \frame{ 36 | \centerline{\textbf{\Huge{课程介绍}}} 37 | } 38 | 39 | %tables 40 | 41 | \section*{大纲} 42 | 43 | \frame{\tableofcontents} 44 | 45 | %introduction 46 | 47 | \section{简介} 48 | \subsection{定义} 49 | 50 | \frame{\frametitle{简介} 51 | 计算智能也被称作“软计算”,是根据自然界生物体系的原理和规律, 52 | 模仿设计出具有记忆、学习、适应等特性的求解算法的总称。 53 | 54 | ~ 55 | 56 | 这些算法通过计算机模拟,再现了生物的某些智能行为。 57 | } 58 | 59 | \subsection{历史} 60 | 61 | \frame{\frametitle{历史} 62 | \begin{enumerate} 63 | \item<1-> 早期探索阶段(1995年之前)。 64 | \item<2-> 成型阶段(1995年-2000年)。 65 | \item<3-> 发展阶段(2001年-2010年)。 66 | \item<4-> 成熟阶段(2011年以来)。 67 | \end{enumerate} 68 | } 69 | 70 | \frame{\frametitle{历史-早期探索} 71 | 这一阶段虽然没有产生具有标志性的生物智能优化算法,但是却为后来的持续快速发展奠定了坚实的基础。 72 | } 73 | 74 | \frame{\frametitle{历史-成型} 75 | 在探索阶段的基础上一些生物智能优化算法开始设计成型。这一时期产生了三个非常经典的生物智能优化算法, 76 | 即由Kennedy和Eberhart于1995年提出的粒子群优化算法、 77 | 由Storn和Price于1997年提出的差分进化算法以及由Shi和Eberhart于1998年提出的基于惯性权重的粒子群优化算法。 78 | 粒子群优化算法是受鸟类的觅食行为的激励而提出的;差分进化算法是受生物进化过程的激励而提出的。 79 | 需要注意的是,这一阶段产生的生物智能优化算法的数量是比较少的。 80 | 但是,这一阶段产生的算法的设计框架一直被沿用至今。 81 | } 82 | 83 | \frame{\frametitle{历史-发展} 84 | 相较于成型阶段,发展阶段所产出的生物智能优化算法的数量显著增加。 85 | 这意味着生物智能优化算法在探索阶段和成型阶段的基础上开始逐渐走向成熟。 86 | 这一阶段也产生了非常具有代表性的算法,比如布谷鸟搜索算法和生物地理学优化算法。 87 | 布谷鸟搜索算法是受一些布谷鸟繁殖行为的启发而提出的; 88 | 生物地理学优化算法是受生物的迁徙现象的激励而提出的。 89 | 布谷鸟搜索算法首次引入了莱维飞行机制来增强全局搜索能力。 90 | 受此启发,现在很多学者已经把莱维飞行作为一种提升生物智能优化算法搜索性能的改进技术。 91 | 生物地理学优化算法得益于其设计的算法结构,具有较强的局部搜索能力。 92 | 因而,生物地理学优化算法经常与其它算法混合以获取更具竞争力的算法, 93 | 比如生物地理学优化算法与灰狼优化算法的混合、 94 | 生物地理学优化算法与布谷鸟搜索算法的混合以及生物地理学优化算法与粒子群优化算法的混合。 95 | } 96 | 97 | \frame{\frametitle{历史-成熟} 98 | 得益于前面三个阶段的积累,自2011年以来,大量的生物智能优化算法相继被学者们提出。 99 | 此外,值得一提是,在这一阶段,学者们采用生物智能优化算法成功地解决了许多不同类型的实际工程问题。 100 | 这表明生物智能优化算法的发展已经进入了较为成熟的阶段。 101 | 这一阶段也产生了许多代表性的生物智能优化算法,比如回溯搜索算法和教与学优化算法。 102 | 回溯搜索算法首次将历史种群用于指导产生下一代种群,并将服从标准正态分布的随机数引入到元启发式算法的设计中; 103 | 而教与学优化算法仅需要必需参数,即种群大小和终止条件,来执行搜索操作。 104 | } 105 | 106 | %methods 107 | 108 | \section{方法} 109 | 110 | \frame{\frametitle{方法} 111 | \begin{itemize} 112 | \item<1-> 进化(演化)算法(遗传算法,协同进化) 113 | \item<2-> 免疫算法 114 | \item<3-> 群集智能(蚁群算法,粒子群算法,飞鸟,鱼群,布谷鸟,蝙蝠,萤火虫) 115 | \item<4-> 模拟退火(源自物理,引出神经网络) 116 | \item<5-> 人工神经网络(连接型,模糊粗糙,记忆型,脉冲耦合,机器学习,深度) 117 | \end{itemize} 118 | } 119 | 120 | \frame{\frametitle{方法} 121 | \begin{itemize} 122 | \item<1-> 专家系统 123 | \item<2-> 生态动力系统 124 | \item<3-> 复杂网络和动力学 125 | \end{itemize} 126 | } 127 | 128 | \frame{\frametitle{方法} 129 | 都是\textsl{仿生算法}。 130 | 131 | ~ 132 | 133 | 其最大特点就是不需要建立问题本身精确的数学模型, 134 | 适合于解决那些因为难以建立有效的形式化模型而用传统人工智能技术 135 | 又难以有效解决甚至无法解决的问题。 136 | } 137 | 138 | \frame{\frametitle{按激励源分类} 139 | \begin{enumerate} 140 | \item<1-> 受生物进化现象激励的智能优化算法。 141 | 这类算法是受自然界中生物的进化过程的启发而提出的,比如经典的差分进化算法和遗传算法; 142 | \item<2-> 受动物的日常行为激励的智能优化算法。 143 | 这类算法是受动物的觅食、围猎以及求偶等行为的启发而提出的,比如粒子群优化算法、布谷鸟搜索算法和灰狼优化算法等等; 144 | \item<3-> 受植物的习性激励的智能优化算法,比如杂草优化算法; 145 | \item<4-> 受人类活动激励的智能优化算法,比如排队搜索算法、人工神经网络算法以及教与学优化算法等; 146 | \item<5-> 受自然界中的物理现象激励的智能优化算法,比如水循环算法、重力搜索算法等。 147 | \end{enumerate} 148 | } 149 | 150 | \frame{\frametitle{按参数类型分类} 151 | \begin{enumerate} 152 | \item<1-> 必需参数。 153 | 指任何一个生物智能优化算法都需要的,其包括种群大小和终止条件(比如最大迭代次数和最大函数迭代次数等)。 154 | 比如教与学优化算法、人工神经网络算法等; 155 | \item<2-> 控制参数。 156 | 指除了必需参数之外反映算法特征的参数,比如差分进化算法中的缩放因子和交叉率、粒子群优化算法中的惯性权重和学习因子、布谷鸟搜索算法中的发现概率等。 157 | \end{enumerate} 158 | } 159 | 160 | \frame{\frametitle{优点} 161 | 162 | {\CJKfamily{zhfs} 163 | 智能计算具有简单、通用、鲁棒性强、适于并行处理的优点, 164 | 使其在并行搜索、联想记忆、模式识别、知识自动获取等方面得到了广泛的应用。 165 | } 166 | 167 | ~ 168 | 169 | 本课目的之一,要能说出各种方法的\textbf{本质}和\textbf{思维共性}。 170 | } 171 | 172 | \frame{\frametitle{缺点} 173 | 174 | 本课目的之一,要能说出每种方法的假设、缺点和适用范围。 175 | } 176 | 177 | \frame{\frametitle{与传统人工智能对比} 178 | 179 | {\CJKfamily{zhhei} 180 | 计算智能有别于传统的符号智能。符号智能是以知识为基础,通过推理进行问题求解,也即传统的人工智能; 181 | 而计算智能则是以数据为基础,通过训练建立联系,进行问题求解。 182 | } 183 | 184 | ~ 185 | 186 | {\CJKfamily{zhkai} 187 | 计算智能是以联接主义为主的思维方式,即研究简单个体如何在简单交互规则指导下, 188 | 构成具有复杂智能行为的高层系统。 189 | } 190 | 191 | ~ 192 | 193 | 数据驱动,研究以相关性为主,因果性为辅。这二者分别对应着定性解释和定量解释。概率统计是其基本工具。 194 | } 195 | 196 | \frame{\frametitle{现状} 197 | 由于过于强调仿生,导致现在的研究针对每一种前人未述及的动物行为就发表大量论文,形成了滥觞。 198 | 在研究范式上停滞不前。 199 | } 200 | 201 | \frame{\frametitle{挑战} 202 | \begin{enumerate} 203 | \item<1-> 无免费午餐理论 204 | \item<2-> 参数设置 205 | \end{enumerate} 206 | } 207 | 208 | \frame{\frametitle{挑战-无免费午餐理论} 209 | 210 | 一个优化算法在一个具体问题上能够获取理想的解决方案, 211 | 并不意味着这一优化算法也能够在其它问题上取得理想的解决方案。 212 | 简而言之,依据“无免费午餐理论”,不存在能够解决所有优化问题的算法。 213 | 也就是说,一个算法在求解一类问题上的成功不能保证这种算法也能在其它类型的问题上得到理想的解决方案。 214 | 不同类型的优化问题通常具有不同的特点,比如变量的多少、目标函数的复杂程度以及约束条件的松紧等, 215 | 这是支持无免费午餐理论的基础。 216 | } 217 | 218 | \frame{\frametitle{挑战-无免费午餐理论} 219 | 例子1: 220 | 221 | ~ 222 | 223 | 综合学习的粒子群优化算法是粒子群优化算法的一个非常经典的改进版本。 224 | 作者采用16个基准测试函数检验所提出的综合学习的粒子群优化算法的性能。 225 | 依据实验结果,当基准测试函数的维度为30时,综合学习的粒子群优化算法优于比较的算法在10个基准测试函数上, 226 | 而在其它6个基准测试函数上,综合学习的粒子群优化算法并没有取得最优的解 227 | (Liang JJ, Qin AK, Suganthan PN, Baskar S. 228 | Comprehensive learning particle swarm optimizer for global optimization of multimodal functions [J]. 229 | IEEE Transactions on Evolutionary Computation, 2006, 10(3): 281–295) 230 | } 231 | 232 | \frame{\frametitle{挑战-无免费午餐理论} 233 | 例子2: 234 | 235 | ~ 236 | 237 | 集成粒子群优化算法是粒子群优化算法的一个改进版本。 238 | 作者用25个基准测试函数检验所提出的集成粒子群优化算法的性能。 239 | 依据实验结果,集成粒子群优化算法在19个基准测试函数上优于粒子群优化算法; 240 | 集成粒子群优化算法在4个基准测试函数上与粒子群优化算法具有相同的性能; 241 | 集成粒子群优化算法在2个基准测试函数上劣于粒子群优化算法 242 | (Lynn N, Suganthan PN. Ensemble particle swarm optimizer [J]. 243 | Applied Soft Computing, 2017, 55: 533–548) 244 | } 245 | 246 | \frame{\frametitle{挑战-参数角度} 247 | 绝大部分群智能优化算法除了必需参数之外,都需要控制参数。 248 | 具有控制参数的群智能优化算法通常会面临一个难题: 249 | 在求解不同特点的优化问题时,如何设置它们的控制参数才能获取最优的解决方案? 250 | 假如控制参数设置得当,这些算法可以得到比较理想的解决方案; 251 | 否则,这些算法可能得到比较糟糕的解决方案。 252 | 显然,控制参数的设置将会制约群智能优化算法的应用潜力。 253 | } 254 | 255 | \frame{\frametitle{挑战-参数角度} 256 | 例子: 257 | 258 | ~ 259 | 260 | 布谷鸟搜索算法中的一个非常重要的参数称为发现概率。 261 | 下图给出了布谷鸟搜索算法在两个不同测试函数上的性能, 262 | 从图(a)可以看到,对于sphere函数,最优的发现概率是0.25; 263 | 对于rosenbrock函数来说,最优的发现概率是0.05。 264 | 265 | \begin{minipage}[b]{\textwidth} 266 | \begin{figure}[hb] 267 | \centering 268 | \includegraphics[width=0.8\textwidth]{0-files/qzn.png} 269 | \end{figure} 270 | \end{minipage} 271 | } 272 | 273 | \frame{\frametitle{挑战} 274 | 从理论角度来说,受“无免费午餐理论”和“参数困境”的激励, 275 | 设计更多简单、高效且仅需要必需参数的群智能优化算法, 276 | 以求解不同类型的优化问题。 277 | } 278 | 279 | \section{应用} 280 | 281 | \frame{\frametitle{应用} 282 | 283 | 自80年代中后期以来计算智能在众多领域的科学家加入下得到了极大的发展。 284 | 285 | ~ 286 | 287 | \textcolor{blue}{看看有哪些应用领域?} 288 | } 289 | 290 | \frame{\frametitle{应用} 291 | 现实生产生活中,优化问题经常碰到,比如电力经济调度问题、机器人路径规划问题、图像分割问题、 292 | 时间序列预测问题等等。 293 | 294 | ~ 295 | 296 | 工程角度看,优化问题就是具有不止一个可行解决方案的问题。 297 | 求解优化问题就是从众多可行方案中选择一个最理想的可行方案。 298 | 这个最理想的解决方案也称为全局最优解决方案。 299 | } 300 | 301 | \section{工作流程} 302 | 303 | \frame{\frametitle{工作流程} 304 | \begin{enumerate} 305 | \item<1-> 清晰地描述问题 306 | \item<2-> 描述数据,识别变量,不变量和隐变量 307 | \item<3-> 探索量之间的关系,借鉴仿生系统,建立数学模型 308 | \item<4-> 数字实验和评价 309 | \item<5-> 对模型进行数学抽象和逻辑分析,给出假设 310 | \item<6-> 给出直观的几何意义和物理意义 311 | \item<7-> 和已有方法进行对比,给出优缺点和适用范围 312 | \item<8-> 撰写报告和论文 313 | \end{enumerate} 314 | } 315 | 316 | \section{课程报告} 317 | 318 | \frame{\frametitle{课程报告} 319 | \begin{enumerate} 320 | \item 每位同学分多次提交报告 321 | \begin{itemize} 322 | \item 对已有的一个方法的深入阐述,有余力的同学可编程实验并评价 323 | \item 探索新的方法 324 | \item 针对科研中的问题,提出计算智能方法并试验 325 | \end{itemize} 326 | \item 提交地址:学在浙大 327 | \end{enumerate} 328 | } 329 | 330 | \frame{\frametitle{报告形式} 331 | \begin{minipage}[t]{\textwidth} 332 | 报告可充分应用图形摘要、漫画等趣味形式,形成既科普,又专业的作品。 333 | \end{minipage} 334 | 335 | \uncover<2->{ 336 | \begin{minipage}[b]{\textwidth} 337 | \begin{figure}[hb] 338 | \centering 339 | \includegraphics[width=0.8\textwidth]{0-files/ant.png} 340 | \end{figure} 341 | \end{minipage} 342 | } 343 | } 344 | 345 | \frame{\frametitle{算法对比} 346 | 347 | \begin{table}[htbp] 348 | \centering 349 | \begin{tabular}{|l|c|r|} 350 | \hline 351 | 算法 & 精度 & 运算时间(s)\\ 352 | \hline 353 | 蚁群 & 90\% & 10.3\\ 354 | \hline 355 | \rowcolor[rgb]{0.0, 0.8, 0.8} 356 | 进化 & \textcolor[rgb]{1.0, 1.0, 0.0}{91.2\%} & 20.6\\ 357 | \hline 358 | 免疫 & 93.1\% & 15.9\\ 359 | \hline 360 | \end{tabular} 361 | \end{table} 362 | } 363 | 364 | \frame{\frametitle{口号} 365 | 366 | \centerline{\textbf{\Huge{\textcolor[rgb]{0.5, 0.5, 0.0}{我们不做数学符号的搬运工!}}}} 367 | } 368 | 369 | \frame{\frametitle{结束} 370 | 371 | \centerline{\textbf{\Huge{结束!}}} 372 | } 373 | 374 | \frame{\frametitle{版权申明} 375 | 376 | 本作品采用知识共享 署名-非商业性使用-禁止演绎 3.0 中国大陆 许可协议进行许可。 377 | 要查看该许可协议,可访问 http://creativecommons.org/licenses/by-nc-nd/3.0/cn/ 或者 378 | 写信到 Creative Commons, PO Box 1866, Mountain View, CA 94042, USA。 379 | } 380 | 381 | \end{document} 382 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 2, June 1991 3 | 4 | Copyright (C) 1989, 1991 Free Software Foundation, Inc., 5 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 6 | Everyone is permitted to copy and distribute verbatim copies 7 | of this license document, but changing it is not allowed. 8 | 9 | Preamble 10 | 11 | The licenses for most software are designed to take away your 12 | freedom to share and change it. By contrast, the GNU General Public 13 | License is intended to guarantee your freedom to share and change free 14 | software--to make sure the software is free for all its users. This 15 | General Public License applies to most of the Free Software 16 | Foundation's software and to any other program whose authors commit to 17 | using it. (Some other Free Software Foundation software is covered by 18 | the GNU Lesser General Public License instead.) You can apply it to 19 | your programs, too. 20 | 21 | When we speak of free software, we are referring to freedom, not 22 | price. Our General Public Licenses are designed to make sure that you 23 | have the freedom to distribute copies of free software (and charge for 24 | this service if you wish), that you receive source code or can get it 25 | if you want it, that you can change the software or use pieces of it 26 | in new free programs; and that you know you can do these things. 27 | 28 | To protect your rights, we need to make restrictions that forbid 29 | anyone to deny you these rights or to ask you to surrender the rights. 30 | These restrictions translate to certain responsibilities for you if you 31 | distribute copies of the software, or if you modify it. 32 | 33 | For example, if you distribute copies of such a program, whether 34 | gratis or for a fee, you must give the recipients all the rights that 35 | you have. You must make sure that they, too, receive or can get the 36 | source code. And you must show them these terms so they know their 37 | rights. 38 | 39 | We protect your rights with two steps: (1) copyright the software, and 40 | (2) offer you this license which gives you legal permission to copy, 41 | distribute and/or modify the software. 42 | 43 | Also, for each author's protection and ours, we want to make certain 44 | that everyone understands that there is no warranty for this free 45 | software. If the software is modified by someone else and passed on, we 46 | want its recipients to know that what they have is not the original, so 47 | that any problems introduced by others will not reflect on the original 48 | authors' reputations. 49 | 50 | Finally, any free program is threatened constantly by software 51 | patents. We wish to avoid the danger that redistributors of a free 52 | program will individually obtain patent licenses, in effect making the 53 | program proprietary. To prevent this, we have made it clear that any 54 | patent must be licensed for everyone's free use or not licensed at all. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | GNU GENERAL PUBLIC LICENSE 60 | TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 61 | 62 | 0. This License applies to any program or other work which contains 63 | a notice placed by the copyright holder saying it may be distributed 64 | under the terms of this General Public License. The "Program", below, 65 | refers to any such program or work, and a "work based on the Program" 66 | means either the Program or any derivative work under copyright law: 67 | that is to say, a work containing the Program or a portion of it, 68 | either verbatim or with modifications and/or translated into another 69 | language. (Hereinafter, translation is included without limitation in 70 | the term "modification".) Each licensee is addressed as "you". 71 | 72 | Activities other than copying, distribution and modification are not 73 | covered by this License; they are outside its scope. The act of 74 | running the Program is not restricted, and the output from the Program 75 | is covered only if its contents constitute a work based on the 76 | Program (independent of having been made by running the Program). 77 | Whether that is true depends on what the Program does. 78 | 79 | 1. You may copy and distribute verbatim copies of the Program's 80 | source code as you receive it, in any medium, provided that you 81 | conspicuously and appropriately publish on each copy an appropriate 82 | copyright notice and disclaimer of warranty; keep intact all the 83 | notices that refer to this License and to the absence of any warranty; 84 | and give any other recipients of the Program a copy of this License 85 | along with the Program. 86 | 87 | You may charge a fee for the physical act of transferring a copy, and 88 | you may at your option offer warranty protection in exchange for a fee. 89 | 90 | 2. You may modify your copy or copies of the Program or any portion 91 | of it, thus forming a work based on the Program, and copy and 92 | distribute such modifications or work under the terms of Section 1 93 | above, provided that you also meet all of these conditions: 94 | 95 | a) You must cause the modified files to carry prominent notices 96 | stating that you changed the files and the date of any change. 97 | 98 | b) You must cause any work that you distribute or publish, that in 99 | whole or in part contains or is derived from the Program or any 100 | part thereof, to be licensed as a whole at no charge to all third 101 | parties under the terms of this License. 102 | 103 | c) If the modified program normally reads commands interactively 104 | when run, you must cause it, when started running for such 105 | interactive use in the most ordinary way, to print or display an 106 | announcement including an appropriate copyright notice and a 107 | notice that there is no warranty (or else, saying that you provide 108 | a warranty) and that users may redistribute the program under 109 | these conditions, and telling the user how to view a copy of this 110 | License. (Exception: if the Program itself is interactive but 111 | does not normally print such an announcement, your work based on 112 | the Program is not required to print an announcement.) 113 | 114 | These requirements apply to the modified work as a whole. If 115 | identifiable sections of that work are not derived from the Program, 116 | and can be reasonably considered independent and separate works in 117 | themselves, then this License, and its terms, do not apply to those 118 | sections when you distribute them as separate works. But when you 119 | distribute the same sections as part of a whole which is a work based 120 | on the Program, the distribution of the whole must be on the terms of 121 | this License, whose permissions for other licensees extend to the 122 | entire whole, and thus to each and every part regardless of who wrote it. 123 | 124 | Thus, it is not the intent of this section to claim rights or contest 125 | your rights to work written entirely by you; rather, the intent is to 126 | exercise the right to control the distribution of derivative or 127 | collective works based on the Program. 128 | 129 | In addition, mere aggregation of another work not based on the Program 130 | with the Program (or with a work based on the Program) on a volume of 131 | a storage or distribution medium does not bring the other work under 132 | the scope of this License. 133 | 134 | 3. You may copy and distribute the Program (or a work based on it, 135 | under Section 2) in object code or executable form under the terms of 136 | Sections 1 and 2 above provided that you also do one of the following: 137 | 138 | a) Accompany it with the complete corresponding machine-readable 139 | source code, which must be distributed under the terms of Sections 140 | 1 and 2 above on a medium customarily used for software interchange; or, 141 | 142 | b) Accompany it with a written offer, valid for at least three 143 | years, to give any third party, for a charge no more than your 144 | cost of physically performing source distribution, a complete 145 | machine-readable copy of the corresponding source code, to be 146 | distributed under the terms of Sections 1 and 2 above on a medium 147 | customarily used for software interchange; or, 148 | 149 | c) Accompany it with the information you received as to the offer 150 | to distribute corresponding source code. (This alternative is 151 | allowed only for noncommercial distribution and only if you 152 | received the program in object code or executable form with such 153 | an offer, in accord with Subsection b above.) 154 | 155 | The source code for a work means the preferred form of the work for 156 | making modifications to it. For an executable work, complete source 157 | code means all the source code for all modules it contains, plus any 158 | associated interface definition files, plus the scripts used to 159 | control compilation and installation of the executable. However, as a 160 | special exception, the source code distributed need not include 161 | anything that is normally distributed (in either source or binary 162 | form) with the major components (compiler, kernel, and so on) of the 163 | operating system on which the executable runs, unless that component 164 | itself accompanies the executable. 165 | 166 | If distribution of executable or object code is made by offering 167 | access to copy from a designated place, then offering equivalent 168 | access to copy the source code from the same place counts as 169 | distribution of the source code, even though third parties are not 170 | compelled to copy the source along with the object code. 171 | 172 | 4. You may not copy, modify, sublicense, or distribute the Program 173 | except as expressly provided under this License. Any attempt 174 | otherwise to copy, modify, sublicense or distribute the Program is 175 | void, and will automatically terminate your rights under this License. 176 | However, parties who have received copies, or rights, from you under 177 | this License will not have their licenses terminated so long as such 178 | parties remain in full compliance. 179 | 180 | 5. You are not required to accept this License, since you have not 181 | signed it. However, nothing else grants you permission to modify or 182 | distribute the Program or its derivative works. These actions are 183 | prohibited by law if you do not accept this License. Therefore, by 184 | modifying or distributing the Program (or any work based on the 185 | Program), you indicate your acceptance of this License to do so, and 186 | all its terms and conditions for copying, distributing or modifying 187 | the Program or works based on it. 188 | 189 | 6. Each time you redistribute the Program (or any work based on the 190 | Program), the recipient automatically receives a license from the 191 | original licensor to copy, distribute or modify the Program subject to 192 | these terms and conditions. You may not impose any further 193 | restrictions on the recipients' exercise of the rights granted herein. 194 | You are not responsible for enforcing compliance by third parties to 195 | this License. 196 | 197 | 7. If, as a consequence of a court judgment or allegation of patent 198 | infringement or for any other reason (not limited to patent issues), 199 | conditions are imposed on you (whether by court order, agreement or 200 | otherwise) that contradict the conditions of this License, they do not 201 | excuse you from the conditions of this License. If you cannot 202 | distribute so as to satisfy simultaneously your obligations under this 203 | License and any other pertinent obligations, then as a consequence you 204 | may not distribute the Program at all. For example, if a patent 205 | license would not permit royalty-free redistribution of the Program by 206 | all those who receive copies directly or indirectly through you, then 207 | the only way you could satisfy both it and this License would be to 208 | refrain entirely from distribution of the Program. 209 | 210 | If any portion of this section is held invalid or unenforceable under 211 | any particular circumstance, the balance of the section is intended to 212 | apply and the section as a whole is intended to apply in other 213 | circumstances. 214 | 215 | It is not the purpose of this section to induce you to infringe any 216 | patents or other property right claims or to contest validity of any 217 | such claims; this section has the sole purpose of protecting the 218 | integrity of the free software distribution system, which is 219 | implemented by public license practices. Many people have made 220 | generous contributions to the wide range of software distributed 221 | through that system in reliance on consistent application of that 222 | system; it is up to the author/donor to decide if he or she is willing 223 | to distribute software through any other system and a licensee cannot 224 | impose that choice. 225 | 226 | This section is intended to make thoroughly clear what is believed to 227 | be a consequence of the rest of this License. 228 | 229 | 8. If the distribution and/or use of the Program is restricted in 230 | certain countries either by patents or by copyrighted interfaces, the 231 | original copyright holder who places the Program under this License 232 | may add an explicit geographical distribution limitation excluding 233 | those countries, so that distribution is permitted only in or among 234 | countries not thus excluded. In such case, this License incorporates 235 | the limitation as if written in the body of this License. 236 | 237 | 9. The Free Software Foundation may publish revised and/or new versions 238 | of the General Public License from time to time. Such new versions will 239 | be similar in spirit to the present version, but may differ in detail to 240 | address new problems or concerns. 241 | 242 | Each version is given a distinguishing version number. If the Program 243 | specifies a version number of this License which applies to it and "any 244 | later version", you have the option of following the terms and conditions 245 | either of that version or of any later version published by the Free 246 | Software Foundation. If the Program does not specify a version number of 247 | this License, you may choose any version ever published by the Free Software 248 | Foundation. 249 | 250 | 10. If you wish to incorporate parts of the Program into other free 251 | programs whose distribution conditions are different, write to the author 252 | to ask for permission. For software which is copyrighted by the Free 253 | Software Foundation, write to the Free Software Foundation; we sometimes 254 | make exceptions for this. Our decision will be guided by the two goals 255 | of preserving the free status of all derivatives of our free software and 256 | of promoting the sharing and reuse of software generally. 257 | 258 | NO WARRANTY 259 | 260 | 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY 261 | FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN 262 | OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES 263 | PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED 264 | OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 265 | MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS 266 | TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE 267 | PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, 268 | REPAIR OR CORRECTION. 269 | 270 | 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 271 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR 272 | REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, 273 | INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING 274 | OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED 275 | TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY 276 | YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER 277 | PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE 278 | POSSIBILITY OF SUCH DAMAGES. 279 | 280 | END OF TERMS AND CONDITIONS 281 | 282 | How to Apply These Terms to Your New Programs 283 | 284 | If you develop a new program, and you want it to be of the greatest 285 | possible use to the public, the best way to achieve this is to make it 286 | free software which everyone can redistribute and change under these terms. 287 | 288 | To do so, attach the following notices to the program. It is safest 289 | to attach them to the start of each source file to most effectively 290 | convey the exclusion of warranty; and each file should have at least 291 | the "copyright" line and a pointer to where the full notice is found. 292 | 293 | {description} 294 | Copyright (C) {year} {fullname} 295 | 296 | This program is free software; you can redistribute it and/or modify 297 | it under the terms of the GNU General Public License as published by 298 | the Free Software Foundation; either version 2 of the License, or 299 | (at your option) any later version. 300 | 301 | This program is distributed in the hope that it will be useful, 302 | but WITHOUT ANY WARRANTY; without even the implied warranty of 303 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 304 | GNU General Public License for more details. 305 | 306 | You should have received a copy of the GNU General Public License along 307 | with this program; if not, write to the Free Software Foundation, Inc., 308 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. 309 | 310 | Also add information on how to contact you by electronic and paper mail. 311 | 312 | If the program is interactive, make it output a short notice like this 313 | when it starts in an interactive mode: 314 | 315 | Gnomovision version 69, Copyright (C) year name of author 316 | Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 317 | This is free software, and you are welcome to redistribute it 318 | under certain conditions; type `show c' for details. 319 | 320 | The hypothetical commands `show w' and `show c' should show the appropriate 321 | parts of the General Public License. Of course, the commands you use may 322 | be called something other than `show w' and `show c'; they could even be 323 | mouse-clicks or menu items--whatever suits your program. 324 | 325 | You should also get your employer (if you work as a programmer) or your 326 | school, if any, to sign a "copyright disclaimer" for the program, if 327 | necessary. Here is a sample; alter the names: 328 | 329 | Yoyodyne, Inc., hereby disclaims all copyright interest in the program 330 | `Gnomovision' (which makes passes at compilers) written by James Hacker. 331 | 332 | {signature of Ty Coon}, 1 April 1989 333 | Ty Coon, President of Vice 334 | 335 | This General Public License does not permit incorporating your program into 336 | proprietary programs. If your program is a subroutine library, you may 337 | consider it more useful to permit linking proprietary applications with the 338 | library. If this is what you want to do, use the GNU Lesser General 339 | Public License instead of this License. 340 | --------------------------------------------------------------------------------