Theory and Principled Methods for the Design of Metaheuristics

This edited book puts together theoretical and principled methods for incorporating problem knowledge in evolutionary algorithms and other metaheuristics, and for designing and adapting them to the specific problems' features. The first part of the book suggests a line of theoretical enquiry aimed at deriving new search algorithms from "scratch" based on well-formulated assumptions on the class of problems considered. The second part of the book is more pragmatic and presents principled methods to adapt and design search operators to the problem at hand. It also demonstrates how statistical methods and other theory-driven tools can use experimental data to enhance the performance of existing metaheuristics in a systematic way. This includes analysing and simplifying metahueirstics, tuning parameters and building algorithms' portfolio.

Dr. Yossi Borenstein is the head of risk analytics at the company VisualDNA; he previously held a position at the University of Hertfordshire, and he received his PhD from the University of Essex; his research interests include data analysis, information retrieval, stochastic optimization, artificial intelligence, and evolutionary computation.

Dr. Alberto Moraglio is a Postdoctoral Research Fellow in the School of Computer Science at the University of Birmingham; he previously held a position at the University of Coimbra, and he received his PhD from the University of Essex; his research focus is the theory of evolutionary computation.