Multiobjective Problem Solving from Nature

Multiobjective evolutionary algorithms (MOEAs) and multiobjective problem solving have become important topics of research in the evolutionary computation community over the past 10 years. This is an advanced text aimed at researchers and practitioners in the area of search and optimization. The book focuses on how MOEAs and related techniques can be used to solve problems, particularly in the disciplines of science and engineering. Contributions by leading researchers deal with the concepts of problem, solution, objective, constraint, utility and preference, and show how these concepts are being investigated in current practice. The book is distinguished from other texts on MOEAs in that it is not primarily about the algorithms, nor specific applications, but about the concepts and processes involved in solving problems using a multiobjective approach. Each chapter contributes to the central, deep concepts and themes of the book.

Includes supplementary material: sn.pub/extras

Inhalt
Introduction: Problem Solving, EC and EMO.- Introduction: Problem Solving, EC and EMO.- Exploiting Multiple Objectives: From Problems to Solutions.- Multiobjective Optimization and Coevolution.- Constrained Optimization via Multiobjective Evolutionary Algorithms.- Tackling Dynamic Problems with Multiobjective Evolutionary Algorithms.- Computational Studies of Peptide and Protein Structure Prediction Problems via Multiobjective Evolutionary Algorithms.- Can Single-Objective Optimization Profit from Multiobjective Optimization?.- Modes of Problem Solving with Multiple Objectives: Implications for Interpreting the Pareto Set and for Decision Making.- Machine Learning with Multiple Objectives.- Multiobjective Supervised Learning.- Reducing Bloat in GP with Multiple Objectives.- Multiobjective GP for Human-Understandable Models: A Practical Application.- Multiobjective Classification Rule Mining.- Multiple Objectives in Design and Engineering.- Innovization: Discovery of Innovative Design Principles Through Multiobjective Evolutionary Optimization.- User-Centric Evolutionary Computing: Melding Human and Machine Capability to Satisfy Multiple Criteria.- Multi-competence Cybernetics: The Study of Multiobjective Artificial Systems and Multi-fitness Natural Systems.- Scaling up Multiobjective Optimization.- Fitness Assignment Methods for Many-Objective Problems.- Modeling Regularity to Improve Scalability of Model-Based Multiobjective Optimization Algorithms.- Objective Set Compression.- On Handling a Large Number of Objectives A Posteriori and During Optimization.