Bob Stern of CRC Press, to whom I am indebted, approached me in late 1999 asking if I was interested in developing a second edition of volume I of the Practical Handbook of Genetic Algorithms. My immediate response was an unequivocal "Yes!" This is the fourth book I have edited in the series and each time I have learned more about GAs and people working in the field. I am proud to be associated with each and every person with whom I have dealt with over the years. Each is dedicated to his or her work, committed to the spread of knowledge and has something of significant value to contribute.
This second edition of the first volume comes a number of years after the publication of the first. The reasons for this new edition arose because of the popularity of the first edition and the need to perform a number of functions for the GA community. These "functions" fall into two main categories: the need to keep practitioners abreast of recent discoveries/learning in the field and to very specifically update some of the best chapters from the first volume.
The book leads off with chapter 0, which is the same chapter as the first edition by Jim Everett on model building, model testing and model fitting. An excellent "How and Why." This chapter offers an excellent lead into the whole area of models and offers some sensible discussion of the use of genetic algorithms, which depends on a clear view of the nature of quantitative model building and testing. It considers the formulation of such models and the various approaches that might be taken to fit model parameters. Available optimization methods are discussed, ranging from analytical methods, through various types of hill-climbing, randomized search and genetic algorithms. A number of examples illustrate that modeling problems do not fall neatly into this clear-cut hierarchy. Consequently, a judicious selection of hybrid methods, selected according to the model context, is preferred to any pure method alone in designing efficient and effective methods for fitting parameters to quantitative models.
Chapter 1 by Roubos and Setnes deals with the automatic design of fuzzy rule-based models and classifiers from data. It is recognized that both accuracy and transparency are of major importance and we seek to keep the rule-based models small and comprehensible. An iterative approach for developing such fuzzy rule-based models is proposed. First, an initial model is derived from the data. Subsequently, a real-coded GA is applied in an iterative fashion, together with a rule-based simplification algorithm to optimize and simplify the model, respectively. The proposed modeling approach is demonstrated for a system identification and a classification problem. Results are compared to other