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.

Скачать книгу бесплатно (pdf, 2.37 Mb)

---

Читать «C++ Neural Networks and Fuzzy Logic»