Stoean C., Stoean R. Support Vector Machines and Evolutionary Algorithms for Classification: Single or Together?

Springer International Publishing Switzerland. 2014. - 129 p.
Indisputably, Support Vector Machines (SVM) and Evolutionary Algorithms (EA) are both established algorithmic techniques and both have their merits and success stories. It appears natural to combine the two, especially in the context of classification. Indeed, many researchers have attempted to bring them together in or or the other way. But if I would be asked who could deliver the most complete coverage of all the important aspects of interaction between SVMs and EAs, together with a thorough introduction into the individual foundations, the authors would be my first choice, the most suitable candidates for this endeavor.
It is now more than ten years ago that I first met Ruxandra, and almost ten years since I first met Catalin, and we have shared a lot of exciting research related and more personal (but not less exciting) moments, and more is yet to come, as I hope. Together,we have experienced some cool scientific successes and also a bitter defeat when somebody had the same striking idea on one aspect of SVMand EA combination and published the paper when we had just generated the first, very encouraging experimental results. The idea was not bad, nonetheless, because the paper we did not write won a best paper award.
Catalin and Ruxandra are experts in SVMs and EAs, and they provide more than an overview over the research on the combination of both with a focus on their own contributions: they also point to interesting interactions that desire even more investigation. And, unsurprisingly, they manage to explain the matter in a way that makes the book very approachable and fascinating for researchers or even students who only know one of the fields, or are completely new to both of them. Bochum, February 2014 Mike Preuss