C++ Neural Networks and Fuzzy Logic [Paperback]

Valluru B. Rao (Author), Hayagriva Rao (Author)

Book Description

Publication Date: October 1995

The extensively revised and updated edition provides a logical and easy-to-follow progression through C++ programming for two of the most popular technologies for artificial intelligence--neural and fuzzy programming. The authors cover theory as well as practical examples, giving programmers a solid foundation as well as working examples with reusable code.

Product Details

• Paperback: 549 pages
• Publisher: M & T Books; 2 Pap/Dsk edition (October 1995)
• Language: English
• ISBN-10: 1558515526
• ISBN-13: 978-1558515529
• Product Dimensions: 9.2 x 7 x 0.8 inches
• Shipping Weight: 2.2 pounds
• Average Customer Review: 2.2 out of 5 stars  See all reviews (25 customer reviews)
• Amazon Best Sellers Rank: #1,764,599 in Books (See Top 100 in Books)

More About the Author

Discover books, learn about writers, read author blogs, and more.

Customer Reviews

Most Helpful Customer Reviews

19 of 19 people found the following review helpful:

1.0 out of 5 stars Poorly written and very confusing, December 21, 1999

By 

Mark D'Urso (USA) - See all my reviews

This review is from: C++ Neural Networks and Fuzzy Logic (Paperback)

To say this book is confusing is an understatement. I am a veteran C++ programmer, with a background in financial application development, looking to expand my skill set. I found this book to be a waste of money.

The chapter on object oriented (OO) programming should have been a warning to me. I remember reading it and thinking, "It's a good thing I'm already familiar with OO, because I would never be able to figure it out from this chapter". I'm not terribly familiar with Neural Networks or Fuzzy Logic and, after reading the first four chapters, I am even more confused than when I started. I think that each topic is getting the same poor treatment that the OO chapter received.

The explanations in this book are useless. Here is an excerpt about how to figure out how many patterns a four node Hopfield network may recall (by the way, the reader is never told what a Hopfield Network is, but is left to infer the key characteristics of a Hopfield Network by comparing the given example to a previous example - which was apparently not a Hopfield network).

"If a pattern P has k, less than 4, bit positions with 0 (and so 4-k bit positions with 1), and if pattern Q is to be orthogonal to P, then Q can have 0 or 1 in those k positions, but it must have only 0 in the rest 4-k positions. Since there are two choices for each of the k positions, there are 2[to the kth power] possible patterns orthogonal to P. this number 2[to the kth power] of patterns includes the pattern with all zeroes. So there really are 2[to the kth power]-1 non-zero patterns orthogonal to P. Some of these 2[to the kth power]-1 patterns are not orthogonal to each other. As an example, P can be the pattern 0 1 0 0, which has k = 3 positions with 0. There are 2[to the 3rd power]-1=7 non-zero patterns orthogonal to 0 1 0 0. Among these are patterns 1 0 1 0 and 1 0 0 1, which are not orthogonal to each other, since their dot product is 1 and not 0." Pg 64

Surely there is a better, and more understandable, way to explain this concept (and every other concept in this book).

I get the feeling that the authors are just going through the motions. They must have created an outline of topics that should be included in a book entitled C++ Neural Networks & Fuzzy Logic but spent little time populating each topic with meaningful information. There seems to be just enough information under each heading to enable someone who already knows the subject to say, "yes that is correct".

This book is a mess.

10 of 10 people found the following review helpful:

1.0 out of 5 stars Bad and code is useless, September 21, 2000

By 

G. Powell (Seattle, WA USA) - See all my reviews
(TOP 1000 REVIEWER)    (REAL NAME)   

This review is from: C++ Neural Networks and Fuzzy Logic (Paperback)

I'm sorry, this book is nearly useless. I'm a vetern C++ programmer and I tried to create a fuzzy logic system from the concepts and code in this book and it didn't work out at all. At first I blamed myself then I found some other fuzzy logic books and realized that no, this book just wasn't able to articulate the concepts in a meaningful way. The Neural net code was even worse. I wanted a cook book, do this, get a simple net, do this get a simple fuzzy logic system, now take what you learned and make a real one. Nope.

They do go over the AI terms and types of Neural nets and I did learn something by reading about it but not enough to justify the price of the book.

9 of 9 people found the following review helpful:

1.0 out of 5 stars Not recommended, April 10, 2004

By A Customer

This review is from: C++ Neural Networks and Fuzzy Logic (Paperback)

Being a professional C++ programmer with a background in Applied Math I didn't like this book at all. It's written in a very annoying way: sometimes it sounds like its author is trying to sell the whole concept of NN and Fuzzy Logic to the reader instead of explaining how, when, and WHY fuzzy logic and neural networks work, how to train them properly, and what their limitations are. One characteristic example: author presents a list of companies using fuzzy logic in real systems but never gives any useful details about these systems.

The book is not for a "mathematician" since it often lacks precision, coherence, mathematical rigor, clarity, ... More often than not, you will find wordy explanations instead of simple formulas

It will displease a "programmer" too. The book's title is "C++ Neural Networks and Fuzzy Logic" so one may expect to find some well-thought and proven design ideas on how to implement NN and FL in C++ as well as a decent C++ library. Instead you will find just an amateurish C++ code (like anybody had any doubts that NN can actually be implemented in C++).

Numerous times author uses this "trick": he introduces new concept, delivers a couple of vague statements about the concept, and promises a better explanation later (in the following chapters, next series, etc).

How do you like this for definition: "STABILITY refers to such convergence that facilitates an end to the iterative process". You can find a lot of such "pearls" in this book.

Don't waste your time... There are better books

FYI: the book includes just a floppy disk instead of CD.