Neural networks and fuzzy systems. A dynamical systems approach to machine intelligence. With two floppy disks.

*(English)*Zbl 0755.94024
Prentice-Hall International Editions. Englewood Cliffs, NJ: Prentice- Hall, Inc. xxviii, 449 p. (1992).

The author has written a textbook with an engineering viewpoint combining the areas of Neural Networks and Fuzzy Systems. The introductory material requires only elementary calculus, linear algebra and elementary probability theory. However, more advanced topics use material from signal processing, random processes and control theory.

The book contains two sections. The first section, consisting of six chapters, develops neural network theory and neural dynamical systems. The second section treats fuzzy systems.

Chapter 1 introduces the ideas on neural networks and fuzzy systems developed in the text. In Chapter 2, the idea of neurons as signal functions is developed and the notation used in the book is presented. The model of how the neuron’s membrane changes with time is treated in Chapter 3. Some probability theory and random processes are reviewed in Chapter 4. Also, unsupervised synoptic learning is discussed. In Chapter 5, supervised synoptic learning is covered. Global equilibrium and stability theorems for feedback and feedforward neural networks are established in Chapter 6.

Chapter 7 presents a geometric theory of fuzzy systems and some relations between probability theory and fuzzy systems. A method of combining fuzzy systems and neural networks to form adaptive fuzzy systems is discussed in Chapter 8. In Chapter 9, a comparison between neural and adaptive fuzzy control systems for backing up a truck is presented. Adaptive fuzzy control systems are applied to image processing in Chapter 10. A comparison between fuzzy control systems and Kalman-filter target- tracking is given in Chapter 11.

There are problems and a references list at the end of each of the first eight chapters. The appendix describes the use of the software on two floppy disks accompanying the book.

The book contains two sections. The first section, consisting of six chapters, develops neural network theory and neural dynamical systems. The second section treats fuzzy systems.

Chapter 1 introduces the ideas on neural networks and fuzzy systems developed in the text. In Chapter 2, the idea of neurons as signal functions is developed and the notation used in the book is presented. The model of how the neuron’s membrane changes with time is treated in Chapter 3. Some probability theory and random processes are reviewed in Chapter 4. Also, unsupervised synoptic learning is discussed. In Chapter 5, supervised synoptic learning is covered. Global equilibrium and stability theorems for feedback and feedforward neural networks are established in Chapter 6.

Chapter 7 presents a geometric theory of fuzzy systems and some relations between probability theory and fuzzy systems. A method of combining fuzzy systems and neural networks to form adaptive fuzzy systems is discussed in Chapter 8. In Chapter 9, a comparison between neural and adaptive fuzzy control systems for backing up a truck is presented. Adaptive fuzzy control systems are applied to image processing in Chapter 10. A comparison between fuzzy control systems and Kalman-filter target- tracking is given in Chapter 11.

There are problems and a references list at the end of each of the first eight chapters. The appendix describes the use of the software on two floppy disks accompanying the book.

Reviewer: D.P.Brown (Carbondale)

##### MSC:

94C99 | Circuits, networks |

92B20 | Neural networks for/in biological studies, artificial life and related topics |

93C42 | Fuzzy control/observation systems |

93C95 | Application models in control theory |

94D05 | Fuzzy sets and logic (in connection with information, communication, or circuits theory) |