A. E. Eiben

Parameter control in evolutionary algorithms

The issue of controlling values of various parameters of an evolutionary algorithm is one of the most important and promising areas of research in evolutionary computation: it has a potential of adjusting the algorithm to the problem while solving the problem. In the paper we: 1) revise the terminology, which is unclear and confusing, thereby providing a classification of such control mechanisms, and 2) survey various forms of control which have been studied by the evolutionary computation community in recent years. Our classification covers the major forms of parameter control in evolutionary computation and suggests some directions for further research.

Parameter Tuning for Configuring and Analyzing Evolutionary Algorithms

In this paper we present a conceptual framework for parameter tuning, provide a survey of tuning methods, and discuss related methodological issues. The framework is based on a three-tier hierarchy of a problem, an evolutionary algorithm (EA), and a tuner. Furthermore, we distinguish problem instances, parameters, and EA performance measures as major factors, and discuss how tuning can be directed to algorithm performance and/or robustness. For the survey part we establish dierent taxonomies to categorize tuning methods and review existing work. Finally, we elaborate on how tuning can improve methodology by facilitating well-funded experimental comparisons and algorithm analysis.

Adaptation in evolutionary computation: a survey

Adaptation of parameters and operators is one of the most important and promising areas of research in evolutionary computation; it tunes the algorithm to the problem while solving the problem. In this paper we develop a classification of adaptation on the basis of the mechanisms used, and the level at which adaptation operates within the evolutionary algorithm. The classification covers all forms of adaptation in evolutionary computation and suggests further research.

Genetic algorithms with multi-parent recombination

We investigate genetic algorithms where more than two parents are involved in the recombination operation. We introduce two multi-parent recombination mechanisms: gene scanning and diagonal crossover that generalize uniform, respecively n-point crossovers. In this paper we concentrate on the gene scanning mechanism and we perform extensive tests to observe the effect of different numbers of parents on the performance of the GA. We consider different problem types, such as numerical optimization, constrained optimization (TSP) and constraint satisfaction (graph coloring). The experiments show that 2-parent recombination is inferior on the classical DeJong functions. For the other problems the results are not conclusive, in some cases 2 parents are optimal, while in some others more parents are better.

Comparing parameter tuning methods for evolutionary algorithms

Tuning the parameters of an evolutionary algorithm (EA) to a given problem at hand is essential for good algorithm performance. Optimizing parameter values is, however, a non-trivial problem, beyond the limits of human problem solving.In this light it is odd that no parameter tuning algorithms are used widely in evolutionary computing. This paper is meant to be stepping stone towards a better practice by discussing the most important issues related to tuning EA parameters, describing a number of existing tuning methods, and presenting a modest experimental comparison among them. The paper is concluded by suggestions for future research - hopefully inspiring fellow researchers for further work.

Global Convergence of Genetic Algorithms: A Markov Chain Analysis

In this paper we are trying to make a step towards a concise theory of genetic algorithms (GAs) and simulated annealing (SA). First, we set up an abstract stochastic algorithm for treating combinatorial optimization problems. This algorithm generalizes and unifies genetic algorithms and simulated annealing, such that any GA or SA algorithm at hand is an instance of our abstract algorithm. Secondly, we define the evolution belonging to the abstract algorithm as a Markov chain and find conditions implying that the evolution finds an optimum with probability 1. The results obtained can be applied when designing the components of a genetic algorithm.

On Evolutionary Exploration and Exploitation

Exploration and exploitation are the two cornerstones of problem solving by search. The common opinion about evolutionary algorithms is that they explore the search space by the (genetic) search operators, while exploitation is done by selection. This opinion is, however, questionable. In this paper we give a survey of different operators, review existing viewpoints on exploration and exploitation, and point out some discrepancies between and problems with current views.

Graph Coloring with Adaptive Evolutionary Algorithms

This paper presents the results of an experimental investigation on solving graph coloring problems with Evolutionary Algorithms (EAs). After testing different algorithm variants we conclude that the best option is an asexual EA using order-based representation and an adaptation mechanism that periodically changes the fitness function during the evolution. This adaptive EA is general, using no domain specific knowledge, except, of course, from the decoder (fitness function). We compare this adaptive EA to a powerful traditional graph coloring technique DSatur and the Grouping Genetic Algorithm (GGA) on a wide range of problem instances with different size, topology and edge density. The results show that the adaptive EA is superior to the Grouping (GA) and outperforms DSatur on the hardest problem instances. Furthermore, it scales up better with the problem size than the other two algorithms and indicates a linear computational complexity.

Parameter Control in Evolutionary Algorithms: Trends and Challenges

More than a decade after the first extensive overview on parameter control, we revisit the field and present a survey of the state-of-the-art. We briefly summarize the development of the field and discuss existing work related to each major parameter or component of an evolutionary algorithm. Based on this overview, we observe trends in the area, identify some (methodological) shortcomings, and give recommendations for future research.

Evolutionary algorithms with on-the-fly population size adjustment

In this paper we evaluate on-the-fly population (re)sizing mechanisms for evolutionary algorithms (EAs). Evaluation is done by an experimental comparison, where the contestants are various existing methods and a new mechanism, introduced here. These comparisons consider EA performance in terms of success rate, speed, and solution quality, measured on a variety of fitness landscapes. These landscapes are created by a generator that allows for gradual tuning of their characteristics. Our test suite covers a wide span of landscapes ranging from a smooth one-peak landscape to a rugged 1000-peak one. The experiments show that the population (re)sizing mechanisms exhibit significant differences in speed, measured by the number of fitness evaluations to a solution and the best EAs with adaptive population resizing outperform the traditional genetic algorithm (GA) by a large margin.