AI Techniques for Game Programming (Premier Press Game Development) [Paperback]

Mat Buckland (Author)

Book Description

ISBN-10: 193184108X | ISBN-13: 978-1931841085 | Publication Date: October 14, 2002 | Edition: 1

"AI Techniques for Game Programming" takes the difficult topics of genetic algorithms and neural networks and explains them in plain English. Gone are the tortuous mathematic equations and abstract examples to be found in other books. Each chapter takes readers through the theory a step at a time, explaining clearly how they can incorporate each technique into their own games. After a whirlwind tour of Windows programming, readers will learn how to use genetic algorithms for optimization, path-finding, and evolving control sequences for their game agents. Coverage of neural network basics quickly advances to evolving neural motion controllers for their game agents and applying neural networks to obstacle avoidance and map exploration. Backpropagation and pattern recognition is also explained. By the end of the book, readers will know how to train a network to recognize mouse gestures and how to use state-of-the-art techniques for creating neural networks with dynamic topologies.

Editorial Reviews

Review

"Game programming is without a doubt the most intellectually challenging field of computer science in the world. However, we would be fooling ourselves if we said that we are 'serious' people! Writing (and reading) a game programming book, should be an exciting adventure for both the author and the reader." Andre LaMothe, Series Editor

About the Author

Mat Buckland studied Computer Science at London University, then spent many years as a Risk Management Consultant. He went on to work for a developer producing games for Gremlin Software. He now works as a freelance programmer and AI consultant. Mat has been interested in evolutionary computing and AI in general since he first read about these techniques back in the early '80s. He is the author of the ai-junkie.com Web Site, which provides tutorials and advice on evolutionary algorithms.

Product Details

• Paperback: 480 pages
• Publisher: Course Technology PTR; 1 edition (October 14, 2002)
• Language: English
• ISBN-10: 193184108X
• ISBN-13: 978-1931841085
• Product Dimensions: 9.3 x 7.4 x 1 inches
• Shipping Weight: 8.5 ounces (View shipping rates and policies)
• Average Customer Review: 4.3 out of 5 stars  See all reviews (22 customer reviews)
• Amazon Best Sellers Rank: #579,507 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

29 of 30 people found the following review helpful:

5.0 out of 5 stars Not just for games, April 30, 2003

By 

J. Wiest "whafrog" (St. Paul, MN) - See all my reviews
(REAL NAME)   

This review is from: AI Techniques for Game Programming (Premier Press Game Development) (Paperback)

Beauty is in the eye of the beholder, so let me tell you what I was looking for: a non-academic AI book with practical examples for implementing genetic algorithms and neural nets. I don't have a math degree (though I understand basic algebra and remember a bit of trig), and I learn best by doing hand's on project, not by theorizing.

This book fits the bill perfectly. It is well written, humorous, clear and patient. The examples are interesting enough that you can see how they would be useful for solving other problems, not necessarily game related.

One caveat: if you don't have a decent intro to basic Win32 API programming, get Charles Petzold's Programming Win32 book and get busy. Yes, the first two chapters of AI Techniques are a Win32 refresher (which was good for me because I last wrote Win32 3 years ago...I now do Java only), but I'm pretty sure if you haven't seen it before you'll be lost.

BTW, just to give you an idea how clear the concepts were presented, I've recoded Chapters 3 and 5 as properly OOPed (MVC, etc) Java applications. Just MHO, but Java is a much better platform for this stuff. ...

13 of 14 people found the following review helpful:

5.0 out of 5 stars Buckland DE-jargonises the field of AI, October 1, 2004

By 

Cormac Mac Galloglai "An Ri Ard" (Galway, Ireland) - See all my reviews
(REAL NAME)   

This review is from: AI Techniques for Game Programming (Premier Press Game Development) (Paperback)

This book IS my Masters thesis. I built a multi layered combative system driven wholly by neural networks evolved using genetic algorithms. The learning of agents was unsupervised and they existed collaboratively and adversarily. If all of this made no sense, dont worry! Read this book and it will.
The book covers these techniques (except for multi layered architectures) to a level understood by anyone with a basic knowledge of C++. It totally demystifies NNs and GAs. Other books on these subjects actually put you off the entire concepts of AI by feeding you fear and confusion.
I have completed the book and have read it numerous times. It is going to be invaluable for my development of PHd stuff - his writing introduces areas for potential research.
Im would definately buy other books from this author. I hope he covers other areas of AI soon and puts 'em into print.
Be jaysus, tis mighty I tell ya!!!

10 of 11 people found the following review helpful:

5.0 out of 5 stars Great Stuff, November 8, 2002

By 

Frank Z (USA) - See all my reviews

This review is from: AI Techniques for Game Programming (Premier Press Game Development) (Paperback)

After reading (and enjoying) the author's web tutorials some time ago I was looking forward to receiving this book - and it hasn't disappointed.

The author has managed to squeeze in a pile of information about GAs and neural nets and yet managed to keep the math down to a minimum - which for a moron like me is especially good news! The source code is kept simple and is very easy to follow.

The example programs that accompany each chapter illustrate each technique very well, and more to the point have given me many ideas to try out with my own projects. The mouse gesture recognition example in particular is a great way of teaching backpropagation, something I had completely failed to understand until I read this book.