Natural Computing Series. Berlin: Springer (ISBN 978-3-642-16543-6/hbk; 978-3-642-16544-3/ebook). xii, 216 p. EUR 79.95/net; SFR 115.00; £ 72.00; $ 99.00 (2010).

Biologically inspired algorithms such as evolutionary algorithms and ant colony optimization have found numerous applications for solving problems from computational biology, engineering, logistics, and telecommunications. Many problems arising in these application domains belong to the field of combinatorial optimization. Bio-inspired algorithms have achieved success when applied to such problems in recent years. In contrast to many successful applications of bio-inspired algorithms, the theoretical foundation of these algorithms lags far behind their practical success. This is mainly due to the fact that these algorithms make use of random decisions in different modules. This leads to stochastic processes that are hard to analyze. This book treats bio-inspired computing methods as stochastic algorithms and presents rigorous results on their runtime behavior. The book is meant to give researchers a state-of-the-art presentation of theoretical results on bio-inspired computing methods in the context of combinatorial optimization. It can be used as basic material for courses on bio-inspired computing that are meant for graduate students and advanced undergraduates. The book is organized in three parts. It starts with a general introduction to bio-inspired algorithms and their computational complexity. Later on, different methods that have been developed in recent years, are presented in a comprehensive manner. Then, some of the major results that have been achieved in the field of single-objective optimization are presented. Different problems such as minimum spanning trees, maximum matchings, and the computation of shortest paths are considered. After these studies, classical multi-objective optimization topics such as the classical computation of multi-objective minimum spanning trees are studied and it is shown that multi-objective approaches can lead to provably better algorithms for classical single-objective problems. The book is recommended as basic material for a course on theoretical aspects of bio-inspired computing.