on 5 June 2003
This is not a good book.
The authors have invented a new, biologically inspired optimisation algorithm (Particle Swarm Optimisation). Like all algorithms it is fundamentally quite simple; once you get your head around it. Like all search algorithms it is good at some things and poor at others.
Unfortunately the book is filled with irrelevant details. It is not a book on Swarm Intelligence; it is a book on the authors algorithm - so it is of little use to people seeking to build social simulations in any other way. On the other hand it does not ask rigorous questions of the algorithm, so it is of limited use to people interested in the technique.
The book tries to pitch at a 'popular science' reader, but the authors find it difficult to present a clear argument without vacillating between patronising simplicity and unexplained jargon randomly introduced.
There are much better alternatives (the Santa Fe Institute book, for example) if you are genuinely interested in 'Swarm Intelligence'.
This book is about how a complex system can be constructed from simple processes. It first described the flocks social behavior model. (Flocks, boids etc. are used quite frequently in movie/tv animations to simulate a group of fish etc.) Then follows evolutionary computation theory (including a detailed description of genetic algorithms), human psychology, collective intelligence, and their application/implications. Since I'm doing a project in artificial evolution of boids (Bird-roids), I find this book very helpful in giving me ideas of how the boids should behave in a socially more realistic way. However this book is mainly theory and doesn't include much implementing details of different system it described (I learnt GA implementing details from the other book 'Genetic Algorithms in search optimisation and machine learning' and various webpages). Overall, a very good book for somebody researching into the area of distributed artificial intelligence.
on 3 April 2008
Very thorough, First takes you through the "thinking behind" Swarm intelligence, almost philosophical. Well balanced, the models start to get more complicated and technical as you progress through the latter chapters. Has some other good models such as AxelRods ACM built in, and with this alone gives/demonstrated four different experiments (with actual results) .....
After this book i think you will be well versed in the technical and more informed aspects of Swarm intelligence. Immerses the reader to give well grounded principles to which further research can be done ... after all you will never find ONE BOOK (Bible) On swarm inteligence since every-one builds there own types but what one needs is a foundation an understanding technical as well as more philosophical in order to has the tools to progress.