### 基本信息

- 原书名：Computational Intelligence: Concepts to Implementations
- 原出版社： Morgan Kaufmann

### 编辑推荐

本书是计算智能领域的经典著作，第一作者是著名的群体智能算法——粒子群优化算法的提出者。书中系统地讨论了计算智能的理论、技术及其应用，包括神经网络、模糊系统和演化计算等内容，比较全面地反映了计算智能研究和应用的最新进展，并提出了一种行之有效的思考和使用计算智能的方法。 本书不仅学术水平较高，而且理论结合实际，很具实用价值。不但有丰富的案例研究和习题，而且提供了教辅和C／C++代码(源代码可以在图灵网站本书网页免费注册下载)，非常适合作为高校教材使用。

### 内容简介

### 目录

Foundations

chapter two

Computational Intelligence

chapter three

Evolutionary Computation Concepts and Paradigms

chapter four

Evolutionary Computation Implementations

chapter five

Neural Network Concepts and Paradigms

chapter six

Neural Network Implementations

chapter seven

Fuzzy Systems Concepts and Paradigms

chapter eight

Fuzzy Systems Implementations

chapter nine

Computational Intelligence Implementations

chapter ten

Performance Metrics

### 前言

At a system prototype level, computational intelligence (CI) tools are capableof yielding results in a relatively short time. For instance, the implementation of aconventional expert system often takes one to three years and requires the activeparticipation of a “knowledge engineer” to build the knowledge and rule bases.In contrast, computational intelligence system solutions can often be prototypedin a few weeks to a few months and are implemented using available engineeringand computational resources. Indeed, computational intelligence tools are capableof being applied in many instances by “domain experts” rather than solely by“computer gurus.”

This means that biomedical engineers, for example, can solve problems inbiomedical engineering without relying on outside computer science expertise suchas that required to build knowledge bases for classical expert systems. Furthermore,innovative ways to combine CI tools are cropping up every day. For example, toolshave been developed that incorporate knowledge elements with neural networks,fuzzy logic, and evolutionary computing theory. Such tools are able to solve quicklyclassification and clustering problems that would be extremely time consumingusing other techniques.

The concepts, paradigms, algorithms, and implementation of computationalintelligence and its constituent methodologies—evolutionary computation, neuralnetworks, and fuzzy logic—are the focus of this book. In addition, we emphasizepractical applications throughout, that is, how to apply the concepts, paradigms,algorithms, and implementations discussed to practical problems in engineeringand computer science. This emphasis culminates in the real-world casestudies in a final chapter, which are available on this book’s web site athttp:// www.computelligence.org/issue/CICI/CICI.html.

Computational intelligence is closely related to the field called “soft computing.”There is, in fact, a significant overlap. According to Lotfi Zadeh (1998), the inventorof fuzzy logic and one of the leading proponents of soft computing:

Soft computing is not a single methodology. Rather, it is a consortium of computingmethodologies which collectively provide a foundation for the conception, designand deployment of intelligent systems.At this juncture, the principal members of softcomputing are fuzzy logic (FL), neurocomputing (NC), genetic computing (GC),and probabilistic computing (PC), with the last subsuming evidential reasoning,belief networks, chaotic systems, and parts of machine learning theory. In contrast totraditional hard computing, soft computing is tolerant of imprecision, uncertaintyand partial truth. The guiding principle of soft computing is: “exploit the tolerancefor imprecision, uncertainty and partial truth to achieve tractability, robustness, lowsolution cost and better rapport with reality.”

Zadeh also believes that soft computing is serving as the foundation for the emergingfield of computational intelligence, and that “In this perspective, the differencebetween traditional AI [artificial intelligence] and computational intelligence is thatAI is based on hard computing whereas CI is based on soft computing” (Zadeh1994).We believe that soft computing is a large subset of computational intelligence.We heartily agree with him when he says, “Hybrid intelligent systems are definitelythe wave of the future” (Zadeh 1994).

Some of the material in this book is adapted from Computational IntelligencePC Tools by Eberhart, Dobbins, and Simpson (Academic Press 1996). The extensiverewrite and reorganization of that material reflect the change in our perception ofcomputational intelligence that has occurred over the years. That change is reflectedin an increased emphasis on evolutionary computation as providing a foundationfor CI. It also features significant recent developments in particle swarm optimizationand other evolutionary computation tools.

The primary intended audience for Computational Intelligence: Concepts toImplementations comprises researchers and graduate students with engineering orcomputer science backgrounds and those with a special interest in computationalintelligence and/or systemadaptation. One of the authors [RE] has taught aCI introductorycourse for several years; the material in this book was developed to supportthat course. Other audiences include researchers in fields such as cognitive scienceand the physical sciences and those who are motivated to learn about computationalintelligence via self-study. We assume this book’s users understand the basic conceptsof classical (two-valued) logic, classical set theory, and elementary probabilitytheory. We also assume that readers have a familiarity with computers and a verybasic familiarity with calculus. Knowledge of a computer language such as Java, C,or Visual BASIC is very helpful but not required.

The implementation chapters frequently refer to and list portions of computercode. In Chapters 4 and 6 we use the most common general-purpose, proceduralprogramming language, C, to implement the evolutionary algorithms and the artificialneural networks. Data structures, routines, and finite state machines are usedextensively in the C programming. In Chapters 8 and 9, reflecting programming languageevolution trends, we use an object-oriented programming language instead ofthe procedural programming language C to implement the fuzzy systems and evolutionaryfuzzy systems. There are a variety of object-oriented languages, such as C++,Java, and C#.We use C++ here primarily because it can be looked at as an extensionof the C language.

Organization of the Book

This book is divided into twelve chapters. Chapters 1 and 2 lay the groundwork forthe topic, introducing computational intelligence and its foundations. The next portionof the book includes the “backbone” chapters on the three main constituents ofCI: evolutionary computation, neural networks, and fuzzy logic, in that order. Thisorder provides an initial focus on evolutionary computation, which is presented asproviding a foundation for development of computational intelligence tools involvingneural networks and fuzzy logic. For instance, when we discuss neural networks,we see how evolutionary computation can be used to evolve the weights and structureof feedforward neural networks, and with fuzzy logic, we examine evolutionarycomputation applications to tools built using fuzzy logic. In other words, the evolutionarycomputation theme pervades this book. Within each backbone chapter,we discuss the histories of computational intelligence, evolutionary computation,neural networks, and fuzzy logic.

We follow each backbone chapter with a chapter discussing implementation andexamples. Each one contains a section on implementation considerations thataddresses features frequently incorporated into these implementations, which featureswe chose and why we chose them, and the guidelines to using them, as wellas interactions among them. The implementation chapters are intended to providereaders with the insight to clearly understand “canned,” commercially packagedsoftware applications and to enable amore thorough understanding of software andhardware implementation issues for CI paradigms.Each chapter ends with exercises.

Chapters’ ContentsChapter 1, Foundations, defines terms used throughout the book and briefly reviewsbiological and behavioral motivations for the constituent methodologies of computationalintelligence. This is followed by a brief review of the major application areasfor each methodology, as well as of CI. The chapter concludes with a review of majorcomputational intelligence application areas.

Chapter 2, Computational Intelligence, launches directly into the core subjectof this book. We first review the concepts of adaptation and self-organization, keyto our view of computational intelligence. Then we summarize the brief history ofthe CI field, viewing it from the perspectives of other researchers. This leads us intoa discussion of the relationships among the three major components and how theycooperate and/or are integrated into a computational intelligence system.We presentour definition of computational intelligence, supported by diagrams that place itinto context.

Chapter 3, Evolutionary Computation: Concepts and Paradigms, has beenadapted from the Evolutionary Computation Theory and Paradigms chapter inSwarm Intelligence (Kennedy, Eberhart, and Shi 2001) with updates and augmentations,including recent developments in particle swarm optimization and otherevolutionary computation approaches. After reviewing the history of evolutionarycomputation and giving an overview of the field, we discuss its main paradigms:genetic algorithms, evolutionary programming, evolution strategies, genetic programming,and particle swarm optimization.

Chapter 4, EvolutionaryComputation Implementations, discusses factors to considerwhen implementing evolutionary computation paradigms and presents twoimplementation examples: a canonical genetic algorithm and a real-valued particleswarm that can be run in single-swarm or multiswarm configurations.

Chapter 5, Neural Network Concepts and Paradigms, first briefly presents anoverview of the history of neural networks, then examines what they are and whythey are useful. A discussion of neural network components and terminology follows,with a review of neural network topologies. A more detailed look at neuralnetwork learning and recall comes next, focusing on three of the most common neuralnetwork paradigms: back-propagation, learning vector quantization, and selforganizingfeature map networks. These networks represent the two basic learningtypes: supervised learning (back-propagation) and unsupervised learning (learningvector quantization and self-organizing feature maps).We also briefly discuss hybridnetworks and recurrent networks. Finally, considerations of preprocessing and postprocessingare evaluated.

Chapter 6, Neural Network Implementations, discusses factors to consider whenimplementing artificial neural networks and presents four implementation examples:back-propagation, learning vector quantization, self-organizing feature maps,and evolutionary neural networks.

Chapter 7, Fuzzy Systems Concepts and Paradigms, leads off with a brief reviewof the history of the field, followed by an examination of fuzzy sets and fuzzy logic,the concepts of fuzzy sets, and approximate reasoning. We stress the differencesbetween fuzzy logic and probability, and we present both Mamdani and Takagi–Sugeno–Kang approaches to the design and analysis of fuzzy systems. The chapterconcludes with a look at some design considerations and special topics related tofuzzy systems.

### 媒体评论

- Professor Shun-ichi Amari

RIKEN Brain Science Institute

Japan

Computational Intelligence describes a large, diverse, and evolving field of theories and techniques, all inspired in one way or the other by nature. The three pillars of CI--neural networks, fuzzy systems, and evolutionary computation--along with their many variants, interact in meaningful ways to solve very complex problems. This book is an excellent introduction to the field, greatly suited for an advanced undergraduate/beginning graduate student course, or for an interested scientist or engineer. The authors guide the reader in an easy-flowing way through the history and foundational mathematics toward practical implementation ora few fun-damental problem-solving systems in each area. In the fuzzy set chapters, they picked the most common application tool, fuzzy rule-based systems, even mixing evolutionary design into the implementation. This book is an excellent choice on its own but, as in my case, will form the foundation for our advanced graduate courses in the CI disciplines. ..

- Professor James M. Keller

University of Missouri-Columbia

This excellent new book by Eberhart and Shi asserts that computational intelligence rests on a foundation of evolutionary computation. This refreshing view has set the book apart from other books on Cl. It has an emphasis on practical applications and computational tools, which are very useful and important for further development of the computational intelligence field. I am delighted that I have a copy of this book. ...

- Professor Xin Yao

The Centre of Excellence for Research in

Computational Intelligence and Applications

The University of Birmingham, Edgbaston

Birmingham, United Kingdom