《模式识别导论》教学大纲
课程代码 | 045101151 |
课程名称 | 模式识别导论 |
英文名称 | Introduction to Pattern Recognition |
课程类别 | 选修课 |
课程性质 | 选修 |
学时 | 总学时：40 实验学时：8 实习学时：0 其他学时：0 |
学分 | 2.5 |
开课学期 | 第六学期 |
开课单位 | 计算机科学与工程学院 |
适用专业 | 计算机科学与技术 |
授课语言 | 中文 |
先修课程 | 无 |
课程对毕业要求的支撑 | №1.（工程知识）培养学生熟练掌握英语，掌握扎实的计算机科学与技术专业基本原理、方法和手段等方面的基础知识用于解决复杂工程问题，并通过计算机系统分析、建模和计算等方面的先进方法，为将所学基础知识应用到计算机科学与技术研发和工程实践做好准备。 №2.（问题分析）培养学生能够创造性地利用计算机科学基本原理解决计算机领域遇到的问题。 №3.（设计/开发解决方案）能够设计针对计算机工程复杂问题的解决方案，设计满足特定需求的计算机软硬件系统，并能够在设计环节中体现创新意识，考虑社会、健康、安全、法律、文化以及环境等因素。 №4.（研究） 培养学生具备计算机系统相关知识并对计算机工程复杂问题进行研究，具有计算机系统研发基本能力、具备问题分析和建模的能力，具有系统级的认知能力和实践能力，掌握自底向上和自顶向下的问题分析方法。 №5.（使用现代工具）能够针对计算机工程复杂问题，开发、选择与使用恰当的技术、资源、现代工程工具和信息技术工具。 №6.（工程与社会）能够基于计算机工程相关背景知识进行合理分析，评价计算机工程实践中的复杂问题解决方案对社会、健康、安全、法律以及文化的影响，并理解应承担的责任。 №7.（环境和可持续发展）能够理解和评价针对计算机工程复杂问题的工程实践对环境、社会可持续发展的影响。 №8.（职业规范）具有人文社会科学素养、社会责任感，能够在工程实践中理解并遵守工程职业道德和规范，履行责任。 №9.（个人和团队）能够在多学科背景下的团队中承担个体、团队成员以及负责人的角色。 №10.（沟通）能够就计算机工程复杂问题与全球业界同行及社会公众进行有效沟通和交流，包括撰写报告和设计文稿、陈述发言、清晰表达或回应指令。并具备良好的国际视野，能够在跨文化背景下进行沟通和交流。 №11.（项目管理）理解并掌握计算机工程管理原理与经济决策方法，并能在多学科环境中应用。 №12.（终身学习）学生能够胜任研究性工作，可继续深造攻读硕士、博士，并具备终身学习的能力。 |
课程目标 | 1：能够理解模式识别问题的基本概念和模型。 2：能够理解分类系统的基本设计思路，模型和相关处理过程。 3：能够应用分类器模型解决实际模式识别问题。 |
课程简介 | 模式识别算法和模型的基础包括基于统计的和结构化的方法。用于模式表征的数据结构，特征发现与选择，参数与非参数分类，监督与无监督学习等技术用于解决小规模样本识别问题。在计算机视觉领域的相关应用包括图像分类，目标检测，行为识别等。课程要求学生具有微积分，线性代数和概率论基础知识，以及与科学计算相关的编程经验。 |
教学内容与学时分配 | （一）模式识别简介 3学时 模式识别举例, 模式识别系统与设计, 分类器评价。 （二）贝叶斯决策理论与极大似然 4学时 贝叶斯规则, 损失函数, 判别函数, 贝叶斯误差, 极大似然参数估计。 （三）线性判别 3学时 线性判别函数，一般线性判别函数，梯度下降，误差平方和函数，最小二乘。 （四）神经网络 6学时 神经网络简介，多层感知机，反向传播算法，正则化，径向基函数网络，实用技术。 （五）支持向量机 3学时 最大间隔，凸优化问题，对偶问题，支持向量，核映射。 （六）决策树 4学时 决策树生成算法，信息熵，数据类型，节点分裂规则，修剪技术。 （七）提升方法 3学时 分类器集成，自适应提升方法，理论特性，应用举例。 （八）无监督学习 3学时 参数与无参数方法，极大似然估计，聚类算法类型，K均值聚类算法，模糊C均值聚类算法，多层聚类算法。 （九）主成分分析 3学时 基本概念，主成分分析算法，算法特点，相关理论，应用举例。 |
实验教学（包括上机学时、实验学时、实践学时） | （一）MATLAB及相关工具箱介绍 2学时 （二）决策森林模型实现及应用 4学时 （三）K均值聚类算法实现及应用 2学时 |
教学方法 | 课堂教学，实验教学 |
考核方式 | 平时表现 40% 期末考试 60% |
教材及参考书 | Richard O. Duda, Peter E. Hart, David G. Stork, “Pattern Classification, Second Edition”, John Wiley & sons, Inc. |
制定人及制定时间 | 吴斯 2019年5月8日 |
“Introduction to Pattern Recognition” Syllabus
Course Code | |
Course Title | Introduction to Pattern Recognition |
Course Category | Elective Courses |
Course Nature | Elective Course |
Class Hours | 40 |
Credits | 2.5 |
Semester | 6 |
Institute | School of Computer Science and Engineering |
ProgramOriented | Computer Science and Technology |
Teaching Language | Chinese |
Prerequisites | None |
Student Outcomes (Special Training Ability) | №1. Engineering Knowledge: An ability to apply knowledge of English, solid knowledge of professional basic principles, methods and means of computer science and technology for solving complex engineering problems, to well prepare the required knowledge applied to the computer science and technology research & development and engineering practice through computer systems analysis, modeling and calculation and any other aspects of the advanced approach. №2. Problem Analysis: An ability to creatively use the basic principles of computer science to solve the problems encountered in the computer field. №3. Design / Development Solutions: An ability to design solutions for computer engineering complex problems, to design computer hardware and software systems that meet with specific requirements, and to embody innovation awareness in the design process and take into account social, health, safety, cultural and environmental factors. №4. Research: An ability to develop computer system-related knowledge and research computer engineering complex issues, to develop the basic capacity of computer systems research & development, systematic cognitive and practice, master the Bottom-up and top-down problem analysis methods. №5. Applying Modern Tools: An ability to develop, select and use appropriate technologies, resources, modern engineering tools and information technology tools for complex computer engineering issues. №6. Engineering and Society: An ability to conduct a reasonable analysis and evaluation of the impact of the solutions of complex problems in computer engineering practice to the social, health, safety, legal and cultural based on computer engineering related background knowledge, and understand the obligation of taking responsibility. №7. Environment and Sustainable Development: An ability to understand and evaluate the impact of solutions of complex engineering problems in environmental and societal contexts and demonstrate knowledge of and need for sustainable development. №8. Professional Standards: An understanding of humanity science and social responsibility, being able to understand and abide by professional ethics and standards responsibly in engineering practice. №9. Individual and Teams: An ability to function effectively as an individual, and as a member or leader in diverse teams and in multi-disciplinary settings. №10. Communication: An ability to communicate effectively on complex computer engineering problems with the engineering community and with society at large, such as being able to comprehend and write effective reports and design documentation, make effective presentations, give and receive clear instructions, and communicate in cross-cultural contexts with international perspective. №11. Project Management: Demonstrate knowledge and understanding of computer engineering management principles and methods of economic decision-making, to function in multidisciplinary environments. №12. Lifelong Learning: An ability to be qualified for research work and can continue their studies for master and doctor degrees, and have the ability for life-long learning. |
Course Objectives | 1: To understand the basic concepts and formulation of pattern recognition problems. 2: To understand the basic design, functions and operations of classification systems. 3: To learn how to apply machine learning and classification techniques in solving real life pattern recognition problems. |
Course Description | Foundations of pattern recognition algorithms and machines, including statistical and structural methods. Data structures for pattern representation, feature discovery and selection, parametric and non-parametric classification, supervised and unsupervised learning, and small sample-size problems. Applications in the computer vision field, such as image classification, object detection, event recognition, etc. It is assumed the students have a working knowledge of calculus, linear algebra, and probability theory. It is also assumed the students have some programming experience in a scientific computing environment. |
Teaching Content and Class Hours Distribution |
Pattern recognition examples, systems and design cycle, and classifier evaluation.
Bayes rule, loss function, discriminant function, Bayes error, and maximum likelihood parameter estimation.
Linear discriminant function, generalized linear discriminant function, gradient descent, sum-of-squared-error function, and least-mean-squared.
Introduction to artificial neural networks, multiple layer perceptron, back-propagation, regularization, radial basis functions, and practical techniques.
Maximum margin, convex optimization, dual optimization problem, support vectors, and kernel trick.
Learning algorithm, entropy, attribute types, splitting rule, and pruning.
Classifier ensemble, adaptive boosting algorithm, theoretical property, and applications.
Parametric and non-parametric methods, maximum likelihood estimates, clustering, K-means clustering algorithm, fuzzy C-means clustering algorithm, and agglomerative clustering algorithm.
Basis concept, PCA algorithm, algorithm characteristics, PCA theory, and applications. |
Experimental Teaching |
|
Teaching Method | Lecture and laboratory experiment |
Examination Method | Usual achievement 40% Final exam 60% |
Teaching Materials and Reference Books | Richard O. Duda, Peter E. Hart, David G. Stork, “Pattern Classification, Second Edition”, John Wiley & sons, Inc. |
Prepared by Whom and When | Si Wu May 8, 2019 |
《模式识别导论》实验教学大纲
课程代码 | 045101151 |
课程名称 | 模式识别导论 |
英文名称 | Introduction to Pattern Recognition |
课程类别 | 选修课 |
课程性质 | 选修 |
学时 | 总学时：40 实验：8 实习：其他： |
学分 | 2.5 |
开课学期 | 第六学期 |
开课单位 | 计算机科学与工程学院 |
适用专业 | 计算机科学与技术 |
授课语言 | 中文 |
先修课程 | 无 |
毕业要求（专业培养能力） | №1.（工程知识）培养学生熟练掌握英语，掌握扎实的计算机科学与技术专业基本原理、方法和手段等方面的基础知识用于解决复杂工程问题，并通过计算机系统分析、建模和计算等方面的先进方法，为将所学基础知识应用到计算机科学与技术研发和工程实践做好准备。 №2.（问题分析）培养学生能够创造性地利用计算机科学基本原理解决计算机领域遇到的问题。 №3.（设计/开发解决方案）能够设计针对计算机工程复杂问题的解决方案，设计满足特定需求的计算机软硬件系统，并能够在设计环节中体现创新意识，考虑社会、健康、安全、法律、文化以及环境等因素。 №4.（研究） 培养学生具备计算机系统相关知识并对计算机工程复杂问题进行研究，具有计算机系统研发基本能力、具备问题分析和建模的能力，具有系统级的认知能力和实践能力，掌握自底向上和自顶向下的问题分析方法。 №5.（使用现代工具）能够针对计算机工程复杂问题，开发、选择与使用恰当的技术、资源、现代工程工具和信息技术工具。 №6.（工程与社会）能够基于计算机工程相关背景知识进行合理分析，评价计算机工程实践中的复杂问题解决方案对社会、健康、安全、法律以及文化的影响，并理解应承担的责任。 №7.（环境和可持续发展）能够理解和评价针对计算机工程复杂问题的工程实践对环境、社会可持续发展的影响。 №8.（职业规范）具有人文社会科学素养、社会责任感，能够在工程实践中理解并遵守工程职业道德和规范，履行责任。 №9.（个人和团队）能够在多学科背景下的团队中承担个体、团队成员以及负责人的角色。 №10.（沟通）能够就计算机工程复杂问题与全球业界同行及社会公众进行有效沟通和交流，包括撰写报告和设计文稿、陈述发言、清晰表达或回应指令。并具备良好的国际视野，能够在跨文化背景下进行沟通和交流。 №11.（项目管理）理解并掌握计算机工程管理原理与经济决策方法，并能在多学科环境中应用。 №12.（终身学习）学生能够胜任研究性工作，可继续深造攻读硕士、博士，并具备终身学习的能力。 |
课程培养学生的能力（教学目标） |
|
课程简介 | 模式识别实验教学包括介绍MATLAB编程软件和相关工具箱， 编程实现决策森林模型并应用在分类问题，编程实现K均值聚类算法并应用在聚类问题。 |
主要仪器设备与软件 | 电脑，MATLAB |
实验报告 | 英文实验报告应包括问题介绍，方法，实验及分析，和总结4个部分。 |
考核方式 | 对实验报告进行评分 |
教材、实验指导书及教学参考书目 | Richard O. Duda, Peter E. Hart, David G. Stork, “Pattern Classification, Second Edition”, John Wiley & sons, Inc. |
制定人及发布时间 | 吴斯 2019年5月8日 |
《模式识别导论》实验教学内容与学时分配
实验项目编号 | 实验项目名称 | 实验学时 | 实验内容提要 | 实验类型 | 实验要求 | 每组人数 | 主要仪器设备与软件 |
1 | MATLAB | 2 | 介绍MATLAB软件及相关工具箱 | 演示性 | 必做 | 1 | 电脑，MATLAB |
2 | 分类 | 4 | 编程实现决策森林模型并应用于分类问题 | 验证性 | 必做 | 1 | 电脑，MATLAB |
3 | 聚类 | 2 | 编程实现K均值聚类算法并应用于聚类问题 | 验证性 | 必做 | 1 | 电脑，MATLAB |
“Introduction to Pattern Recognition” Syllabus
Course Code | |
Course Title | Introduction to Pattern Recognition |
Course Category | Elective Courses |
Course Nature | Elective Course |
Class Hours | 40 |
Credits | 2.5 |
Semester | 6 |
Institute | School of Computer Science and Engineering |
Program Oriented | Computer Science and Technology |
Teaching Language | Chinese |
Prerequisites | None |
Student Outcomes (Special Training Ability) | №1. Engineering Knowledge: An ability to apply knowledge of English, solid knowledge of professional basic principles, methods and means of computer science and technology for solving complex engineering problems, to well prepare the required knowledge applied to the computer science and technology research & development and engineering practice through computer systems analysis, modeling and calculation and any other aspects of the advanced approach. №2. Problem Analysis: An ability to creatively use the basic principles of computer science to solve the problems encountered in the computer field. №3. Design / Development Solutions: An ability to design solutions for computer engineering complex problems, to design computer hardware and software systems that meet with specific requirements, and to embody innovation awareness in the design process and take into account social, health, safety, cultural and environmental factors. №4. Research: An ability to develop computer system-related knowledge and research computer engineering complex issues, to develop the basic capacity of computer systems research & development, systematic cognitive and practice, master the Bottom-up and top-down problem analysis methods. №5. Applying Modern Tools: An ability to develop, select and use appropriate technologies, resources, modern engineering tools and information technology tools for complex computer engineering issues. №6. Engineering and Society: An ability to conduct a reasonable analysis and evaluation of the impact of the solutions of complex problems in computer engineering practice to the social, health, safety, legal and cultural based on computer engineering related background knowledge, and understand the obligation of taking responsibility. №7. Environment and Sustainable Development: An ability to understand and evaluate the impact of solutions of complex engineering problems in environmental and societal contexts and demonstrate knowledge of and need for sustainable development. №8. Professional Standards: An understanding of humanity science and social responsibility, being able to understand and abide by professional ethics and standards responsibly in engineering practice. №9. Individual and Teams: An ability to function effectively as an individual, and as a member or leader in diverse teams and in multi-disciplinary settings. №10. Communication: An ability to communicate effectively on complex computer engineering problems with the engineering community and with society at large, such as being able to comprehend and write effective reports and design documentation, make effective presentations, give and receive clear instructions, and communicate in cross-cultural contexts with international perspective. №11. Project Management: Demonstrate knowledge and understanding of computer engineering management principles and methods of economic decision-making, to function in multidisciplinary environments. №12. Lifelong Learning: An ability to be qualified for research work and can continue their studies for master and doctor degrees, and have the ability for life-long learning. |
Teaching Objectives |
|
Course Description | The laboratory experiments include the introduction to the software MATLAB and related toolboxes, implementation of the decision forest model for classification tasks, and implementation of the K-means clustering algorithm for clustering tasks. |
Instruments and Equipments | Computer, MATLAB |
Experiment Report | Experiment reports should include the following four parts: the introduction to the task, proposed approach, experimental results and discussion, and conclusion. |
Assessment | Grade experiment reports |
Teaching Materials and Reference Books | Richard O. Duda, Peter E. Hart, David G. Stork, “Pattern Classification, Second Edition”, John Wiley & sons, Inc. |
Prepared by Whom and When | Si Wu May 8, 2019 |
CourseTitle” Experimental Teaching Arrangements
No. | Experiment Item | Class Hours | Content Summary | Category | Requirements | Number of StudentsEach Group | Instruments, Equipments and Software |
1 | MATLAB | 2 | Introduce the software MATLAB and related toolboxes. | Demonstration | Compulsory | 1 | Computer, MATLAB |
2 | Classification | 4 | Implement the decision forest model and apply it to a classification task. | Verification | Compulsory | 1 | Computer, MATLAB |
3 | Clustering | 2 | Implement the K-means clustering algorithm and apply it to a clustering task. | Verification | Compulsory | 1 | Computer, MATLAB |
专业课程思政建设内容
序号 | 课程名称 | 任课教师 | 职称 | 学院 | 育人目标 | 教学特色 | 预期成效 |
1 | 模式识别导论 | 吴斯 | 副教授 | 计算机科学与工程学院 | 1. 结合十九大确立的国家战略规划及本院的人才培养目标，培养学生坚定正确的政治方向、强烈的社会责任感； | 讲述近现代以来机器学习方法的发展历程，紧密结合专业知识启蒙与思政教育，培养正确的专业伦理，激发学生将个人理想与国家建设大计紧密结合，传承一代又一代优秀工科领军人物脚踏实地、立足创新的优良作风，传承追求卓越的创造精神和精益求精的品质精神。 | 1.以“模式识别导论”的第一堂课为抓手，实现计算机专业知识教学与立德树人教育的有机融合，引导学生热爱所学专业； |
