Applying Evolutionary Algorithms Successfully: A Guide Gained from Real-world Applications

TL;DR

This paper offers practical guidance for applying evolutionary algorithms effectively in real-world optimization problems, based on 30 years of research and experience with GLEAM and HyGLEAM, emphasizing project management and design strategies.

Contribution

It provides a comprehensive, experience-based framework for novice users to apply EAs successfully, including common pitfalls and project management considerations.

Findings

Effective application depends on understanding problem-specific EA design.

Experience with GLEAM and HyGLEAM informs best practices.

Non-technical factors significantly influence project success.

Abstract

Metaheuristics (MHs) in general and Evolutionary Algorithms (EAs) in particular are well known tools for successful optimization of difficult problems. But when is their application meaningful and how does one approach such a project as a novice? How do you avoid beginner's mistakes or use the design possibilities of a metaheuristic search as efficiently as possible? This paper tries to give answers to these questions based on 30 years of research and application of the Evolutionary Algorithm GLEAM and its memetic extension HyGLEAM. Most of the experience gathered and discussed here can also be applied to the use of other metaheuristics such as ant algorithms or particle swarm optimization. This paper addresses users with basic knowledge of MHs in general and EAs in particular who want to apply them in an optimization project. For this purpose, a number of questions that arise in the course of such a project are addressed. At the end, some non-technical project management issues are discussed, whose importance for project success is often underestimated.