Customer Reviews

Learning and Soft Computing: Support Vector Machines, Neural Networks, and Fuzzy Logic Models (Complex Adaptive Systems)

5 star:		(3)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

What strikes me each time I open this book is Mr Kecman's sense of pedagogy: it is a lesson in the matter. Not only his book delivers the - sometimes complex - techniques in a highly readable manner, but the concepts behind each of the main tools (SVM, NN & FL) he chose to highlight are always brilliantly put in context. One comes out of the reading with more than a set of equations but rather with a clearer picture of the field.
Mr Kecman is - without a doubt - a great teacher.

This effort to deliver a clear message is furthermore underlined through the numerous original figures: if you are like me and feel that a (good) picture speaks more than a thousand words, you will sure appreciate the way the illustrations complement the text and truly help the understanding.

I have read several other books on the subject but if I had to chose one for teaching purposes, this would be the one. I you want to build a better understanding of the field, get this book: it will pay on the long term.

This book is a nice and, I would say, a successful attempt to provide a unified survey of important theoretical and practical machine learning tools: neural networks (NN), support vector machines (SVM) and fuzzy systems (FS).

Book consists of nine chapters, covering SVMs, one- and multi-layer perceptrons and radial-basis function networks, as variants of neural networks, and basics of fuzzy theory. This is followed by interesting case-studies (in financial, control and computer graphic applications) and concluded by basics of optimization theory and an overview of necessary mathematical tools. All the MATLAB programs needed for the simulated experiments are available on the book web site.

Authored by Vojislav Kecman, a prominent researcher in the field of soft computing and previous MIT visiting professor, this book is an excellent material for advanced undergraduate and introductory graduate courses in machine learning applications and soft computing....

This is a very good book. Another reviewer has commented on Vojislav Kecman being an excellent teacher. I whole-heartedly second that opinion. Often times, while reading this book, you will pause with a doubt or question. What you will find surprising is that almost certainly the author has answered that question in the next paragraph. Many times, the author's answers will tally your own answers.

The first chapter of the book (entitled: Learning and Soft Computing: Rationale, Motivations, Needs, Basics) is 119 pages long. It is an essential reading. By the time you finish reading this chapter the things will start falling into place and you will be more motivated and ready to read the remaining chapters. Until you are highly aware of this topic, do not skip this chapter.

A book is made up of a lot of things other than the text that it covers. Does it contain many/any stupid jokes? Is it printed on the highest quality paper? Is the font size good? Is it printed too dense? Is the cover page inviting enough? Are the dimensions/weight of the book correct? On all these counts the book scores high.

Consistent with the subject matter that it covers, this is not an easy book. You will perhaps like to read it with paper and pencil. But if you are willing to spend time with this book, this book will do a lot of good to you. This is a very good book.