Customer Reviews

Efficient and Accurate Parallel Genetic Algorithms (Genetic Algorithms and Evolutionary Computation 1)

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

 

 

Genetic algorithms are easy to parallelize, but they are difficult to control. In a very concise manner, this book presents some theoretical results derived by the author that show how to make parallel genetic algoritms work for many problems and different architectures.

The book has a lot of new theory that is easy to follow and gives recommendations to make parallel genetic algorithms work well in many circumstances. Although the theory makes many simplyfying assumptions, the examples in the book demonstrate that the models are very accurate and the recommendations made in the book seem very reasonable.

I think the previous reviewer got it wrong: It is very clear from the editorial introduction and from the table of contents that this book is about the ANALYSIS of parallel genetic algorithms, not about their implementation. The problem here is that implementing these algorithms is relatively easy, but configuring them is very complicated because there are many parameters. The work of Cantu-Paz---excellently summarized in this little book---is the best analysis available for the different types of parallel genetic algorithms.

I recommend this book wholeheartedly.

One would have thought from the title of the book that it would contain "... Parallel Genetic Algorithms", however it does not. The book covers statistical analyses of algorithms. The reader is meant to find other references that contain algorithms. It is expensive for such a small book. The author talks 'around' the algorithms, without going into the necessary details to write your own parallel genetic algorithms. There are many plots of evolution, so the author must of had access to the algorithms. This book could have been so much more valuable if it contained the algorithms, rather than just talk about them.