本书是《模式识别及Matlab实现》主教材的配套实验与指导，根据主教材各章内容，相应给出了实验的具体步骤和程序代码，包括：贝叶斯决策，概率密度函数的参数估计，非参数判别分类方法，聚类分析，特征提取与选择，模糊模式识别，神经网络在模式识别中的应用，模式识别的工程应用等。

本书是武汉理工大学杨杰和郭志强编写的教材《模式识别及MATLAB 实现》的学习和 实验指导用书，可与教材配套使用，也可单独作为高等学校模式识别课程的教学与学习参考 书，还可作为模式识别领域专业技术人员的参考资料。 模式识别是一门理论和工程应用都发展十分迅速的学科，尤其随着大数据的出现和互联 网+的兴起，模式识别已伴随着人工智能技术渗透到人们生活的方方面面。“模式识别”作 为信息类专业硕士研究生的学位课，主要介绍模式识别的基础知识和基本理论，为进一步研 究模式识别理论和技术打下良好的基础。同时，模式识别也是一门实践性很强的学科，通过 一定量的实验训练，有助于学习者加深理解和巩固所学的基本理论知识，也有助于提高其解 决实际工程问题的能力。 全书分为7 章，每章都按本章知识结构、知识要点和实验指导三部分编写。具体内容包 括贝叶斯决策、参数估计、非参数判别分类法、聚类分析法、特征选择与提取、模糊模式识 别、数字图像处理的基础等，每章实验均给出了实验步骤、MATLAB 代码和实验结果。实验 的内容和训练对模式识别学习者有很大帮助，也为从事模式识别的工程技术人员提供了一定 的指导。 本书第1～4 章由郭志强编写，第5～7 章由杨杰编写，编者指导的研究生王贺、吴紫薇、 林仲康和李博闻等参加了程序调试、插图和校对工作。在编写本书过程中，参阅了大量模式 识别参考书，这里谨向有关作者表示衷心感谢。 由于作者水平有限，书中难免存在疏漏和不当之处，恳请读者批评指正。

第 1 章贝叶斯决策 ·························································································· 1 1.1 知识要点 ····························································································· 1 1.2 实验指导 ····························································································· 7 1.2.1 基于最小错误率的贝叶斯决策 ························································· 7 1.2.2 最小风险判决规则 ······································································· 12 1.2.3 最大似然比判决规则 ···································································· 16 1.2.4 Neyman-Pearsen 判决 ···································································· 21 第2 章参数估计 ···························································································· 25 2.1 知识要点 ···························································································· 25 2.2 实验指导 ···························································································· 30 2.2.1 最大似然估计 ············································································· 30 2.2.2 贝叶斯估计 ················································································ 33 2.2.3 Parzen 窗 ··················································································· 36 2.2.4 N k 近邻估计法 ············································································ 38 第3 章非参数判别分类法 ················································································ 41 3.1 知识要点 ···························································································· 41 3.2 实验指导 ···························································································· 44 3.2.1 两分法 ······················································································ 44 3.2.2 两分法的设计 ············································································· 47 3.2.3 没有不确定区域的两分法 ······························································ 52 3.2.4 广义线性判别函数的设计与实现 ····················································· 56 3.2.5 感知器算法的设计/实现 ································································ 58 3.2.6 两类问题Fisher 准则 ···································································· 62 3.2.7 基于距离的分段线性判别函数 ························································ 68 3.2.8 支持向量机 ················································································ 74 第4 章聚类分析法 ························································································· 80 4.1 知识要点 ··························································································· 81 4.2 实验指导 ··························································································· 84 4.2.1 距离测度 ··················································································· 84 4.2.2 相似测度算法 ············································································· 90 4.2.3 基于匹配测度算法的实现 ······························································ 98 4.2.4 基于类间距离测度方法 ································································ 103 4.2.5 聚类函数准则 ············································································ 106 4.2.6 基于最近邻规则的聚类算法 ·························································· 108 4.2.7 基于最大最小距离聚类算法的实现 ················································· 113 4.2.8 基于K-均值聚类算法实验 ···························································· 116 第5 章特征提取与选择 ·················································································· 124 5.1 知识要点 ·························································································· 124 5.2 实验指导 ·························································································· 128 5.2.1 基于距离的可分性判据 ································································ 128 5.2.2 图像的傅里叶变换二（旋转性质） ················································· 130 5.2.3 基于熵函数的可分性判据 ····························································· 134 5.2.4 利用类均值向量提取特征 ····························································· 136 5.2.5 基于类平均向量中判别信息的最优压缩的实现 ·································· 141 5.2.6 增添特征法 ··············································································· 144 5.2.7 剔减特征法 ··············································································· 148 5.2.8 增l 减r（算法）的设计/实现 ························································ 151 5.2.9 分支定界法（BAB 算法） ···························································· 156 第6 章模糊模式识别 ····················································································· 161 6.1 知识要点 ·························································································· 161 6.2 实验指导 ·························································································· 163 6.2.1 最大隶属度识别法 ······································································ 163 6.2.2 择近原则识别法 ········································································· 167 6.2.3 基于模糊等价关系的聚类算法研究 ················································· 170 第7 章数字图像处理的基础 ··········································································· 179 7.1 知识要点 ·························································································· 179 7.2 实验指导 ·························································································· 181 7.2.1 前馈神经网络感知器的设计实现 ··················································· 181 7.2.2 基于BP 网络的多层感知器 ·························································· 184 7.2.3 自组织特征映射网络的设计/实现 ·················································· 189 7.2.4 径向基神经网络 ········································································ 194 参考文献 ······································································································· 198

教学：主要承担数字图像处理，模式识别，信息传输理论等课程的教学工作。__eol____eol__科研：近几年先后承担了国家自然科学基金面上项目，湖北省科技攻关项目，海南科技厅项目，武汉市科技攻关项目，国家留学基金项目，交通部重点项目和企业合作项目30余项；在国内外学术刊物上发表学术论文60余篇，被SCI和EI收录论文40余篇，出版教材3部。 获武汉理工大学优秀硕士学位论文奖，优秀博士论文奖。__eol__