Nature-Inspired Optimization Algorithms

Abstract

NATURE-INSPIRED OPTIMIZATION ALGORITHMS by Xin-She Yang Elsevier, 2014,263 pp. ISBN: 978-0-12-416743-8Nature-Inspired Optimization Algorithms is a survey of the current research in swarm intelligence. Swarm intelligence algorithms are based on the idea that numerous individual actors operating with simple rules can sometimes be more efficient in solving search and optimization problems than one actor performing a single, all-encompassing algorithm. The original inspiration for swarm intelligence dates back to the late 1990's when artificial intelligence researchers began designing parallel algorithms to mimic the activities of ants gathering food, termites building mounds, and birds flying in formation. Since then the field has crossed over from biological modeling to optimization in physical systems.Nature-Inspired Optimization Algorithms introduces the design and application of the major techniques in the field of swarm intelligence. Included are ant and bee algorithms, bat algorithms, cuckoo search, firefly algorithms, bat algorithms, and flower pollination algorithms. Also discussed are earlier algorithms that grew out of artificial intelligence, including simulated annealing and genetic algorithms. The book contains fifteen chapters, divided roughly into five sections, and two significant appendices. Part of the Elsevier Insights series, this book is also available as an eBook. The print volume, which is reviewed here, does not include color or an index.The first three chapters, comprising the first part of the book, provide an introduction to algorithms, their analysis, and the basic mathematics of random walks and optimization. Included in the first chapter is a discussion of Wolpert and Macready's celebrated No Free Lunch Theorem (NFL) for optimization algorithms. Some of the algorithms that are discussed in detail later in the book are introduced in Chapter 2. Chapter 3 is a thorough discussion of random walks and their relation to optimization algorithms.Three chapters comprising the second part of the book describe the early nature-inspired techniques of simulated annealing, genetic algorithms, and differential evolution. Although not swarm intelligence, these algorithms, inspired by the cooling of metals and biological evolution, provided the first notions that algorithms could model natural processes.Chapters 7-11, the third part of the book, provide the details of several more popular swarm intelligence algorithms. Here we find descriptions and assessments of particle swarm optimization (PSO), firefly algorithms (FA), cuckoo search (CS), bat algorithms (BA), and flower pollination algorithms (FPA). …