(20##) 10 9 8 7 6 5 4 3 2 1 0 <
DesktopWeb FormText   book : Genetic AlgorithmsSat, 26 Mar 2005 15:23:11 GMT # 

in Search Optimization and Machine Learning (Goldberg, 1989). this was a good book on GA (genetic algorithms) with a touch of GBML (genetic based machine learning). it was light on the math and easy to understand. i.e. there were only a couple places i got lost. plus it even had a little bit of humor. overall, i consider it to be the same level of difficulty of the GA book by Mitchell you can find on amazon. the only problem i have with these books is that the samples are too simple / contrived. wish they would have at least one sample that was applied to some real world problem, in which the solution was not a perfect fit, and the author could walk through their reasoning for how they chose the many classifiers. so now i need to actually come up with an idea of something that i want to code using GA ... or move on to reading about GP (genetic programming). which brings up something i'm not clear on. how is GBML different from GP? the way i currently contrast GA and GBML is as such: GA is an off-line search algorithm that you want to converge to some answer, while GBML is an on-line algorithm that can be used to classify. you dont want its population to converge to one answer, but instead you want its population to handle multiple answers. ... or something like that :)