A surrogate-assisted linkage inference approach in genetic algorithms
Tomasz Oliwa, Khaled Rasheed · 2011
Linkage in terms of genetic algorithms is a measure of interdependence of groups of genes. When linkage exists, the fitness contribution of one gene depends on the allele setting of another. Our approach, a surrogate-assisted linkage inference genetic algorithm (SALIGA), is able to detect linkage for real-valued alleles. It uses a perturbation-based method with the aid of fitness surrogates and clustering techniques. Experimental results of linkage inference applied to several synthetic fitness functions are provided. The results demonstrate that SALIGA is able to reliably infer linkage and will correctly group the genes of an individual according to their linkage group membership, while using fewer fitness evaluations through a surrogate model. In addition, our results are augmented with a discussion regarding linkage detection through feature selection.