Book Review: Holger H. Hoos and Thomas Stützle: Stochastic local search: foundations and applications (2005)

Over the last 20 years or so combinatorial optimization has flourished into an area with an almost uncountable number of successful real-world applications. One of the major drivers for developing methods being able to solve even large scale optimization problems was and is the field of metaheuristics. While a proper definition of metaheuristics goes beyond local search based approaches these turned out to be the most widely known and successful approaches. And once local search is combined with randomness one arrives at Stochastic Local Search. The authors describe them as follows: “Many widely known and high-performance local search algorithms make use of randomized choices in generating or selecting candidate solutions for a given combinatorial problem instance. These algorithms are called stochastic local search (SLS) algorithms, and they constitute one of the most successful and widely used approaches for solving hard combinatorial problems.” (Certainly there is a more formal definition given as well.) Naturally, at least once randomized and appended or hybridized by a local search component, these include a wealth of methods such as simulated annealing, iterated local search, greedy randomized adaptive search, variable neighbourhood search, ant colony optimization, among others. So, the title says it all: the book is about SLS and it consists of two parts, foundations and applications. Each part has five chapters following the concept of having a brief outline about what to be expected, the main body of the chapter, some bibliographic remarks, a summary and some exercises. The text itself is interleaved with almost everything you would wish to have, definitions, proper