Customer Reviews

Genetic Algorithms in Search, Optimization, and Machine Learning

5 star:		(13)
4 star:		(2)
3 star:		(2)
2 star:		(2)
1 star:		(0)

 

 

One seldom finds a book as well-written as this one. The underlying mathematics are explained in a very accessible manner, yet with enough rigor to fully explain the "partial schemata" theory which is so important to understanding when and where GenAlgs can be applied. It is the lack of coverage of this theory which causes so much misunderstanding and disappointment in the power of genetic algorithms.

But beyond the background math (which makes up a small part of the book) this is really a tutorial on implementing GenAlgs, and it is an excellent one. The sample code is great, and the implementations are developed throughout the book, allowing the reader to implement simple (but functional) algorithms after reading only the first few chapters, but building to very sophisticated and modern techniques by the end of the book.

A great find.

More than seven years after publication, David Goldberg's clear prose, straightforward code examples, and solid theoretical coverage keeps "the blue book" head-and-shoulders above any other text on this most intriguing of algorithmic directions. This is the book that lifted genetic algorithms from obscurity to one of the most discussed (and misunderstood) of emerging technologies. Goldberg did not invent genetic algorithms (that honor goes to either Nature or John Holland, depending on your personal belief system), but he did make sure that they could be understood by any interested programmer. The source code is in Pascal, which may not be to everyone's taste, but is certainly readable by anyone with a programming background. - Larry O'Brien (Editor, AI Expert Magazine 1990-1994

This book gives a good introduction to genetic algorithms for a general undergraduate audience. However, it is important to note that it does not cover Evolutionary Strategies, an approach to evolutionary computing that I have found quite useful since it is specifically designed for Euclidean space optimization problems where many if not most interesting optimization problems are formulated in (take for example the problem of determining the weights of a neural network that minimizes the network's overall classification error). Nor does it cover evolutionary programming (not to be confused with genetic programming). So after reading this book, I recommend (for the mathematically adventurous) Thomas Back's "Evolutionary Algorithms in Theory and Practice: Evolution Strategies, Evolutionary Programming, Genetic Algorithms"
ISBN: 0195099710

Happy reading and enjoy the fascinating world of evolutionary computation!

I was looking for an automated approach to finding an optimum run sequence through a changeover matrix. The programming examples gave me the elements I needed to experiment and then fine tune the approach for a working search algorithm. I found the book a good companion in my "voyage of discovery".

For me, the book works two levels, the basic pieces to "play with" are presented clearly in chapters 1 and 3, and practical implementation suggestions are spread throughout the text.

By developing programs in Visual Basic, experimenting with search parameters and re-reading sections of this book - I learned something new!

This is the only book I have read about Genetic algorithms, but it seems that it covers the field pretty well.

In the preface it says that it is aimed a beginning graduate students, and since I have a M.Sc. in Computer Science and I just wanted to read it for fun I thought it would be for me. But I found that it uses way to many words to explain very basic things (e.g. almost a page to explain binary numbers) while many of the difficult equations just was presented without proper proof. So the book could have better if it had been cut down to a third and then supplemented with the proper proofs. So if you are a Computer Science graduate I cannot recommend this book. Given the fine books that Addison-Wesley usually publish I was quite disappointed with this one.

But if you are a student in other fields and just want an "intuitive" impression of Genetic Algorithms without the mathematical rigor it is probably good.

Chapter 1: An introduction to genetic algorithms with examples. This chapter is excellent.
Chapter 2: The mathematical theory behind genetic algorithms. This is not done very well since many of the equations isn't proven or explained properly.
Chapter 3: A Pascal program for the sample in chapter 1. This seems unneccesary since any proficient programmer easily could have implemented the program based on the information in chapter 1.
Chapter 4: The history of genetic algorithms and a number of applications all taken from research. Both seem unneccesary.
Chapter 5: An extension of the techniques presented in chapter 1. This is good.
Chapter 6-7: Introduction to machine learning. Is ok.
Chapter 8: A concluding chapter without any real new information.

I bought this book while I was a working professional. It is one of the few textbooks that I have ever read straight through, like a novel. In addition to making everything clear and interesting, the book was even funny at times! I didn't think that was allowed in textbooks. ;-)

I agree with another reviewer who said the book was unnecessarily long. Genetic Algorithms are a great programming tool, and there are some tips and tricks that can help your programs converge faster and more accurately, but this book had a lot of redundant information.

If you are interested in using GA for solution-finding, I doubt you'll find much useful in this book beyond the first chapter or so. Many of the examples later in the book were so specific that I couldn't see how they could be usefully generalized. Really optimizing a GA approach for a specific problem domain takes a fair amount of tuning, and this book won't help much with that.

I think time spent surfing siteseer or other publication sites would be better spent than reading this book.

This is one of the best books I've read for genetic algorithms and AI. I wish all books were like this. It is not pedagogical in its style (unlike Computational Intelligence - Engelbrecht), it is well written, very insightful and slowly takes you into the depths of GA/AI, so it's great to follow. This book contains source codes in Pascal (which is easy to translate to any other language - although you'd want to write your own based on OOP), pseudo codes, examples, and plenty of ways to understand the way GA's work. BUY THIS BOOK and you'll save yourself a lot of sweat and mind boggling wierd explanations from supposedly good authors. I'll never sell this book.

One reader wrote a comment about how this book could be cut in half, and is not suitable for CS majors, my response to that: "I'm a CS major, doing my Ph.D., my professor, my colleagues and almost everyone in the field has a copy of this book, maybe you never got past chapter2 in his book. If you want proof of theorems, there's lots of research papers available, almost all of which refer to Goldberg's book."

This is a great book to begin your journey on Genetic Algorithms (GA). The author is a pioneering authority on the subject and has explained the basics of a GA in a very gentle and easy to understand manner. The book has a great variety of specific but diverse examples, which may not be useful at first glance, but gives an insight to where all the technique has been applied!

However, some aspects of the book perhaps need an edition, like the more recent advances in GA operators, specifics of chromosomal representation schemes, non-linear optimization functions, etc. I have read several, well written books on the subject, but this one has a very distinct and sometimes interesting style of writing! The best would be to quickly read this one to get a fairly good understanding of the basics and then take up a recent book that addresses other aspects like Mitchell's book, for example.

Having said that, I think the book is a great and inspiring start to using genetic algorithms.

At the beginning of the book you will find a nice intro of GA, some comparison with other techniques, a mathematical foundation and some fields of application. Then you get into the real stuff. Coding(in Pascal) of the basic structures of GA and some more sophisticated, like: scaling fitness, inversion, diploid cells etc... there's also a step-by-step simulation to make you comprehend how the GA really works.