人工智能：智能系统指南（英文版·第2版）

　　 人工智能经常被人们认为是计算机科学中的一门高度复杂甚至令人生畏的学科。长期以来人工智能方面的书籍往往包含复杂矩阵代数和微分方程。本书形成于作者多年来给没有多少微积分知识的学生授课时所用的讲义，它假定读者预先没有编程的经验，并说明了智能系统中的大部分基础知识实际上是简单易懂的。
　　 本书目前已经被国际上多所大学(例如，德国的马德堡大学、日本的广岛大学、美国的波士顿大学和罗切斯特理工学院)采用。 如果你正在寻找关于人工智能或智能系统设计课程的浅显易懂的入门级教材，如果你不是计算机科学领域的专业人员，而又正在寻找介绍基于知识系统最新技术发展的自学指南，本书将是最佳选择。
　　 [font color="#ff6600"]本书的主要内容：[/font]
　　 ● 基于规则的专家系统
　　 ● 模糊专家系统
　　 ● 基于框架的专家系统
　　 ● 人工神经网络
　　 ● 进化计算
　　 ● 混合智能系统
　　 ● 知识工程
　　 ● 数据挖掘

preface
prdface to the second edition
acknowledgements
1 introduction to knowledge-based intelligent systems
1.1 intelligent machines, or what machines can do
1.2 the history of artificial intelligence, or from the‘dark ages’to knowledge-based systems
1.3 summary
questions for review
references

2 rule-based expert systems
2.1 introduction, or what is knowledge?
2.2 rules as a knowledge representation technique
2.3 the main players in the expert system development team
2.4 structure of a rule-based expert system
2.5 fundamental characteristics of an expert system
2.6 forward chaining and backward chaining inference techniques
2.7 media advisor: a demonstration rule-based expert system
2.8 conflict resolution
2.9 advantages and disadvantages of rule-based expert systems

　　 'The only way not to succeed is not to try.'
　　 Edward Teller
　　 Another book on artificial intelligence ... I've already seen so many of them. Why should I bother with this one? What makes this book different from the others?
　　 Each year hundreds of books and doctoral theses extend our knowledge of computer, or artificial, intelligence. Expert systems, artificial neural networks, fuzzy systems and evolutionary computation are major technologies used in intelligent systems. Hundreds of tools support these technologies, and thousands of scientific papers continue to push their boundaries. The contents of any chapter in this book can be, and in fact is, the subject of dozens of monographs. However, I wanted to write a book that would explain the basics of intelligent systems, and perhaps even more importantly, eliminate the fear of artificial intelligence.
　　 Most of the literature on artificial intelligence is expressed in the jargon of computer science, and crowded with complex matrix algebra and differential equations. This, of course, gives artificial intelligence an aura of respectability, and until recently kept non-computer scientists at bay. But the situation has changed!
　　 The personal computer has become indispensable in our everyday life. We use it as a typewriter and a calculator, a calendar and a communication system, an interactive database and a decision-support system. And we want more. We want our computers to act intelligently! We see that intelligent systems are rapidly coming out of research laboratories, and we want to use them to our advantage.
　　 What are the principles behind intelligent systems? How are they built? What are intelligent systems useful for? How do we choose the right tool for the job? These questions are answered in this book.
　　 Unlike many books on computer intelligence, this one shows that most ideas behind intelligent systems are wonderfully simple and straightforward. The book is based on lectures given to students who have little knowledge of calculus. And readers do not need to learn a programming language! The material in this book has been extensively tested through several courses taught by the author for the past decade. Typical questions and suggestions from my students influenced the way this book was written.
　　 The book is an introduction to the field of computer intelligence. It covers rule-based expert systems, fuzzy expert systems, frame-based expert systems, artificial neural networks, evolutionary computation, hybrid intelligent systems and knowledge engineering.
　　 In a university setting, this book provides an introductory course for undergraduate students in computer science, computer information systems, and engineering. In the courses I teach, my students develop small rule-based and frame-based expert systems, design a fuzzy system, explore artificial neural networks, and implement a simple problem as a genetic algorithm. They use expert system shells (Leonardo, XpertRule, Level5 Object and Visual Rule Studio), MATLAB Fuzzy Logic Toolbox and MATLAB Neural Network Toolbox. I chose, these tools because they can easily demonstrate the theory being presented. However, the book is not tied to any specific tool; the examples given in the book are easy to implement with different tools.
　　 This book is also suitable as a self-study guide for non-computer science professionals. For them, the book provides access to the state of the art in knowledge-based systems and computational intelligence. In fact, this book is aimed at a large professional audience: engineers and scientists, managers and businessmen, doctors and lawyers everyone who faces challenging problems and cannot solve themby using traditional approaches, everyone who wants to understand the tremendous achievements in computer intelligence. The book will help to develop a practical understanding of what intelligent systems can and cannot do, discover which tools are most relevant for your task and, finally, how to use these tools.
　　 The book consists of nine chapters.
　　 In Chapter 1, we briefly discuss the history of artificial intelligence from the era of great ideas and great expectations in the 1960s to the disillusionment and funding cutbacks in the early 1970s; from the development of the first expert systems such as DENDRAL, MYCIN and PROSPECTOR in the seventies to the maturity of expert system technology and its massive applications in different areas in the 1980s and 1990s; from a simple binary model of neurons proposed in the 1940s to a dramatic resurgence of the field of artificial neural networks in the 1980s; from the introduction of fuzzy set theory and its being ignored by the West in the 1960s to numerous 'fuzzy' consumer products offered by the Japanese in the 1980s and world-wide acceptance of 'soft' computing and computing with words in the 1990s.
　　 In Chapter 2, we present an overview of rule-based expert systems. We briefly discuss what knowledge is, and how experts express their knowledge in the form of production rules. We identify the main players in the expert system development team and show the structure of a rule-based system. We discuss fundamental characteristics of expert systems and note that expert systems can make mistakes. Then we review the forward and backward chaining inference techniques and debate conflict resolution strategies. Finally, the advantages and disadvantages of rule-based expert systems are examined.
　　 In Chapter 3, we present two uncertainty management techniques used in expert systems: Bayesian reasoning and certainty factors. We identify the main sources of uncertain knowledge and briefly review probability theory. We consider the Bayesian method of accumulating evidence and develop a simple expert system based on the Bayesian approach. Then we examine the certainty factors theory (a popular alternative to Bayesian reasoning) and develop an expert system based on evidential reasoning. Finally, we compare Bayesian reasoning and certainty factors, and determine appropriate areas for their applications.
　　 In Chapter 4, we introduce fuzzy logic and discuss the philosophical ideas behind it. We present the concept of fuzzy sets, consider how to represent a fuzzy set in a computer, and examine operations of fuzzy sets. We also define linguistic variables and hedges. Then we present fuzzy rules and explain the main differences between classical and fuzzy rules. We explore two fuzzy inference techniques Mamdani and Sugeno and suggest appropriate areas for their application. Finally, we introduce the main steps in developing a fuzzy expert system, and illustrate the theory through the actual process of building and tuning a fuzzy system.
　　 In Chapter S, we present an overview of frame-based expert systems. We consider the concept of a frame and discuss how to use frames for knowledge representation. We find that inheritance is an essential feature of frame based systems. We examine the application of methods, demons and rules. Finally, we consider the development of a frame-based expert system through an example.
　　 In Chapter 6, we introduce artificial neural networks and discuss the basic ideas behind machine learning. We present the concept of a perceptron as a simple computing element and consider the perceptron learning rule. We explore multilayer neural networks and discuss how to improve the computationai efficiency of the back-propagation learning algorithm. Then we introduce recurrent neural networks, consider the Hopfield network training algorithm and bidirectional associative memory (BAM). Finally, we present self-organising neural networks and explore Hebbian and competitive learning.
　　 In Chapter 7, we present an overview of evolutionary computation. We consider genetic algorithms, evolution strategies and genetic programming. We introduce the main steps in developing a genetic algorithm, discuss why genetic algorithms work, and illustrate the theory through actual applications of genetic algorithms. Then we present a basic concept of evolutionary strategies and determine the differences between evolutionary strategies and genetic algorithms. Finally, we consider genetic programming and its application to real problems.
　　 In Chapter 8, we consider hybrid intelligent systems as a combination of different intelligent technologies. First we introduce a new breed of expert systems, called neural expert systems, which combine neural networks and rulebased expert systems. Then we consider a neum-fuzzy system that is functionally equivalent to the Mamdani fuzzy inference model, and an adaptive neuro-fuzzy inference system (ANFIS), equivalent to the Sugeno fuzzy inference model. Finally, we discuss evolutionary neural networks and fuzzy evolutionary systems.

　　 文艺复兴以降，源远流长的科学精神和逐步形成的学术规范，使西方国家在自然科学的各个领域取得了垄断性的优势；也正是这样的传统，使美国在信息技术发展的六十多年间名家辈出、独领风骚。在商业化的进程中，美国的产业界与教育界越来越紧密地结合，计算机学科中的许多泰山北斗同时身处科研和教学的最前线，由此而产生的经典科学著作，不仅擘划了研究的范畴，还揭橥了学术的源变，既遵循学术规范，又自有学者个性，其价值并不会因年月的流逝而减退。
　　 近年，在全球信息化大潮的推动下，我国的计算机产业发展迅猛，对专业人才的需求日益迫切。这对计算机教育界和出版界都既是机遇，也是挑战；而专业教材的建设在教育战略上显得举足轻重。在我国信息技术发展时间较短、从业人员较少的现状下，美国等发达国家在其计算机科学发展的几十年间积淀的经典教材仍有许多值得借鉴之处。因此，引进一批国外优秀计算机教材将对我国计算机教育事业的发展起积极的推动作用，也是与世界接轨、建设真正的世界一流大学的必由之路。
　　 机械工业出版社华章图文信息有限公司较早意识到"出版要为教育服务"。自1998年开始，华章公司就将工作重点放在了遴选、移译国外优秀教材上。经过几年的不懈努力，我们与PrenticeHall，Addison-Wesley，McGraw-Hill，MorganKaufmann等世界著名出版公司建立了良好的合作关系，从它们现有的数百种教材中甄选出Tanenbaum，Stroustrup，Kernighan，JimGray等大师名家的一批经典作品，以"计算机科学丛书"为总称出版，供读者学习、研究及庋藏。大理石纹理的封面，也正体现了这套丛书的品位和格调。
　　 "计算机科学丛书"的出版工作得到了国内外学者的鼎力襄助，国内的专家不仅提供了中肯的选题指导，还不辞劳苦地担任了翻译和审校的工作；而原书的作者也相当关注其作品在中国的传播，有的还专诚为其书的中译本作序。迄今，"计算机科学丛书"已经出版了近百个品种，这些书籍在读者中树立了良好的口碑，并被许多高校采用为正式教材和参考书籍，为进一步推广与发展打下了坚实的基础。
　　 随着学科建设的初步完善和教材改革的逐渐深化，教育界对国外计算机教材的需求和应用都步入一个新的阶段。为此，华章公司将加大引进教材的力度，在"华章教育"的总规划之下出版三个系列的计算机教材：除"计算机科学丛书"之外，对影印版的教材，则单独开辟出"经典原版书库"；同时，引进全美通行的教学辅导书"Schaum'sOutlines"系列组成"全美经典学习指导系列"。为了保证这三套丛书的权威性，同时也为了更好地为学校和老师们服务，华章公司聘请了中国科学院、北京大学、清华大学、国防科技大学、复旦大学、上海交通大学、南京大学、浙江大学、中国科技大学、哈尔滨工业大学、西安交通大学、中国人民大学、北京航空航天大学、北京邮电大学、中山大学、解放军理工大学、郑州大学、湖北工学院、中国国家信息安全测评认证中心等国内重点大学和科研机构在计算机的各个领域的著名学者组成"专家指导委员会"，为我们提供选题意见和出版监督。
　　 这三套丛书是响应教育部提出的使用外版教材的号召，为国内高校的计算机及相关专业的教学度身订造的。其中许多教材均已为M．I．T.，Stanford，U．C．Berkeley，C．M．U．等世界名牌大学所采用。不仅涵盖了程序设计、数据结构、操作系统、计算机体系结构、数据库、编译原理、软件工程、图形学、通信与网络、离散数学等国内大学计算机专业普遍开设的核心课程，而且各具特色--有的出自语言设计者之手、有的历经三十年而不衰、有的已被全世界的几百所高校采用。在这些圆熟通博的名师大作的指引之下，读者必将在计算机科学的宫殿中由登堂而入室。
　　 权威的作者、经典的教材、一流的译者、严格的审校、精细的编辑，这些因素使我们的图书有了质量的保证，但我们的目标是尽善尽美，而反馈的意见正是我们达到这一终极目标的重要帮助。教材的出版只是我们的后续服务的起点。华章公司欢迎老师和读者对我们的工作提出建议或给予指正，我们的联系方法如下：
　　 电子邮件：hzedu@hzbook．com
　　 联系电话：(010)68995264
　　 联系地址：北京市西城区百万庄南街1号
　　 邮政编码：100037