Evolvable machines. Theory \& practice. (Q1762581)

The articles of this volume will not be indexed individually. Evolutionary algorithms are one of the very promising developments in Computer Science that can solve difficult problems, especially such problems where analytical solutions are not known or their application would be too complex. The basic principle, the adaptation of elements to a given goal function (to a given purpose) via natural selection has been varied, many different approaches are available, however, the basic ideas are still more or less the same. Very often very good modeling skills are required in order to transform a problem in such a way that these algorithms are applicable. The solutions resulting from the application of these ideas are often very ``creative and elegant''. The presented book is an excellent example for the successful application of these methods to the Design of Machines where the concept of a Machine has to be understood in a very broad sense. The first part (4 chapters) deals with the design of mechanical systems (such as cooperating robots, the movement of snake-like robots) and their control systems. Very impressive and interesting results can be seen, it will also be very useful to study the methodology that has been applied. The second part (3 chapters) covers the solution of specific hardware-related problems (optimization of finite automata and Field Programmable Gate Arrays) as well as the design of MicroElectroMechanical Systems (where the mixture of different disciplines is quite obvious). The third and last part (3 chapters) covers the design of special systems (such as Multilayer Perceptrons, Fuzzy Hardware) and even some programming structures. A very interesting book, highly recommendable for scientists interested in evolvable systems, but also for all scientists who have to solve special different problems in various application areas.