Customer Reviews

Genetic Algorithms and Investment Strategies (Wiley Finance)

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

 

 

The author presents a general introduction into ga's , then moves over to investment strategies and presents solutions.

Unfortunately, he does NOT give background information on the really interesting things like string patterns used, crossover and fitness function and the like. Futhermore, more than one third of the book is filled with endless tables whose content the reader understands after the first table. I guess it makes for a larger book.

conclusion: very disappointing.

For a project decision maker trying to evaluate the genetic algorithm approach, this book helps to understand the main concepts. It is a great for anyone who wants ideas to help evaluate the potentials and shortcomings of a GA system without getting mixed up in the math and details. It includes comparisons to chaos theory and neural nets.

For programmers looking implement code, or data administrators looking for the right data feeds, however, this book might be a frustrating tease. The book is full of examples such as mentioning the selection of "10 parameters from a possible 150 macroeconomic variables", but never says what the parameters are. The book is still useful. A little patience with this book might yield some great ideas although the author didn't directly communicate it. Never the less, just one small idea could easily pay for this book many many times over. For the pure IT developer, I highly recommend "Neural Networks in Finance and Investing" by Trippi & Turban instead which includes documented code and examples on a disk.

If you are looking for practical applications of this kind of algoritmicts do not buy this book. The 50% of the book are data tables, it would be better buying a cd-rom with all those data and the book had 200 less pages. A better choice than this book is "Trading on the edge" or "Neural networks for finance".

After I read this book I read the review from 1998. I was quite surprised that the prior reviewer felt that way. The book provides an introduction to developing the data required for testing, a methodology for developing a study that would be useful in investment practice. This is a quality effort. This book will not provide (or promise) the reader a turnkey system for conquering the market, however there is a framework for future research.

This book was written in 1994, before many of the books dealing with Neural Networks came out, and so the terminology will seem unfamiliar. If you are willing to work through these differences (and it is not too hard) then there is a great deal to learn here.

Bauer predicted (in 1994) that Genetic Algorithms would become widely used. Bauer also predicted that much of the development would be done in secret. I have not come across them in the last few years and at very least I would have expected to see them as signals for sale from system developers. Additionally, there are a series of books like this one that should have appeared since 1994. A search of Amazon using Genetic Algorithm as the subject and sorting by publication date returns 133 titles. I reviewed these titles and did not find any further investment focused titles. I will use this book as a starting point for my research.

My next book will be Melanie Mitchell's "Introduction to GAs".