The Science of Breeding and its application to the Breeder Genetic Algorithm(BGA)

Heinz Muhlenbein and Dirk Schlierkamp-Voosen

Reviewer: Nooriyan Poonawala (npoo010@aucklanduni.ac.nz)

References:

1. Muhlenbein, H., Schlierkamp, D. (1993). The science of breeding and the Breeder Genetic Algorithm. Evolutionary Computation, Vol.1 No.4(335-360).

Keywords:

Breeder genetic algorithm, Artificial selection, Heritability, Selection, Recombination, Mutation, Regression.

Related research:

1. Muhlenbein, H., Schlierkamp, D. (1993). Predictive models for Breeder genetic algorithm I. Continuous Parameter Optimization. Evolutionary Computation, Vol.1 No.1(25-49).

Summary:

The breeder genetic algorithm (BGA) described in the paper models artificial selection as performed by human breeders. The science of breeding is based on advanced statistical methods. The main focus of the paper is to establish a connection between the theory behind a genetic algorithm and the science of breeding. The response to selection equation and the concept of heritability is applied to predict the behaviour of the BGA(this includes analysis of selection, mutation and recombination). Mutation and recombination are shown to be complementary search operators. The BGA is also analysed and controlled in this paper, using regression techniques to estimate the heritability arc for general fitness landscapes.

Evaluation:

The paper reviewed was a neat, well structured paper. It has discrete subheadings for all the different areas that were explored, which made it easy to navigate. Although, it utilises several advanced statistical estimates and methods that are not very well described, and hence a sound knowledge of advanced statistics is required to understand the paper fully.