Agentbased Evolutionary Search

This book presents an overview of archiving strategies developed over the last years by the authors that deal with suitable approximations of the sets of optimal and nearly optimal solutions of multi-objective optimization problems by means of stochastic search algorithms. All presented archivers are analyzed with respect to the approximation qualities of the limit archives that they generate and the upper bounds of the archive sizes.

An efficient and adequate constraint-handling technique is a key element in the design of competitive evolutionary algorithms to solve complex optimization problems. This edited book presents a collection of recent advances in nature-inspired techniques for constrained numerical optimization. The book covers six main topics: swarm-intelligence-based approaches, studies in differential evolution, evolutionary multiobjective constrained optimization, hybrid approaches, real-world applications and the recent use of the artificial immune system in constrained optimization. Within the chapters, the reader will find different studies about specialized subjects, such as: special mechanisms to focus the search on the boundaries of the feasible region, the relevance of infeasible solutions in the search process, parameter control in constrained optimization, the combination of mathematical programming techniques and evolutionary algorithms in constrained search spaces and the adaptation of novel nature-inspired algorithms for numerical optimization with constraints.

Evolutionary algorithms are successful biologically inspired meta-heuristics. Their success depends on adequate parameter settings. The question arises: how can evolutionary algorithms learn parameters automatically during the optimization? Evolution strategies gave an answer decades ago: self-adaptation. Their self-adaptive mutation control turned out to be exceptionally successful. But nevertheless self-adaptation has not achieved the attention it deserves.

Intended for researchers and practitioners alike, this book covers carefully selected yet broad topics in optimization, machine learning, and metaheuristics. Written by world-leading academic researchers who are extremely experienced in industrial applications, this self-contained book is the first of its kind that provides comprehensive background knowledge, particularly practical guidelines, and state-of-the-art techniques. New algorithms are carefully explained, further elaborated with pseudocode or flowcharts, and full working source code is made freely available.

Decades of innovations in combinatorial problem solving have produced better and more complex algorithms. These new methods are better since they can solve larger problems and address new application domains. They are also more complex which means that they are hard to reproduce and often harder to fine-tune to the peculiarities of a given problem. This last point has created a paradox where efficient tools are out of reach of practitioners.

Evolutionary algorithms is a class of randomized heuristics inspired by natural evolution. They are applied in many different contexts, in particular in optimization, and analysis of such algorithms has seen tremendous advances in recent years.

Despite decades of work in evolutionary algorithms, there remains a lot of uncertainty as to when it is beneficial or detrimental to use recombination or mutation. This book provides a characterization of the roles that recombination and mutation play in evolutionary algorithms. It integrates prior theoretical work and introduces new theoretical techniques for studying evolutionary algorithms.

It is recognized that many real-world problems can be interpreted and formulated as optimization problems. This feature has fostered the development of research studies aiming to design and implement efficient optimization methods, able to address the increasing complexity of the applications that are intended to be solved. These research studies have mostly followed two main axes.