Adaptation in natural and artificial systems
Adaptation in natural and artificial systems
What Makes a Problem Hard for a Genetic Algorithm? Some Anomalous Results and Their Explanation
Machine Learning - Special issue on genetic algorithms
Niching methods for genetic algorithms
Niching methods for genetic algorithms
Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation
Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation
Problem difficulty for tabu search in job-shop scheduling
Artificial Intelligence
Uniform Crossover in Genetic Algorithms
Proceedings of the 3rd International Conference on Genetic Algorithms
Modeling Building-Block Interdependency
PPSN V Proceedings of the 5th International Conference on Parallel Problem Solving from Nature
Optimization Using Distributed Genetic Algorithms
PPSN I Proceedings of the 1st Workshop on Parallel Problem Solving from Nature
Is the common good? a new perspective developed in genetic algorithms
Is the common good? a new perspective developed in genetic algorithms
Overcoming hierarchical difficulty by hill-climbing the building block structure
Proceedings of the 9th annual conference on Genetic and evolutionary computation
| Hi-index | 0.00 |
The traditional GA theory is pillared on the Building Block Hypothesis (BBH) which states that Genetic Algorithms (GAs) work by discovering, emphasizing and recombining low order schemata in high-quality strings, in a strongly parallel manner. Historically, attempts to capture the topological fitness landscape features which exemplify this intuitively straight-forward process, have been mostly unsuccessful. Population-based recombinative methods had been repeatedly outperformed on the special designed abstract test suites, by different variants of mutation-based algorithms. Departing from the BBH, in this paper we seek to exemplify the utility of crossover from a different point of view, emphasizing the creative potential of the crossover operator. We design a special class of abstract test suites, called Trident functions, which exploits the ability of modern GAs to mix good but significantly different solutions. This approach has been so far neglected as it is widely believed that disruption caused by mating individuals that are too dissimilar may be harmful. We anticipate that hybridizing different designs induces a complex neighborhood structure unattainable by trajectorybased methods which can conceal novel solutions. Empirical results confirm that the proposed class of problems can be solved efficiently only by population-based panmictic recombinative methods, employing diversity maintaining mechanisms.