Genetic Algorithm Encoding Probabilities &Population Size
Dermot W. O’Dwyer, Eugene J. OBrien · WIT transactions on information and communication technologies · 1970
Genetic algorithms offer a potentially powerful tool for tackling optimisation problems which cannot be formulated as mathematical programming problems. This paper explores why success in the application of genetic algorithms has not been universal. The paper first categorises optimisation problems and identifies features of optimisation problems which identify the problems as being suitable for solution using genetic algorithms. The paper shows how choice of coding system affects convergence and gives guidance on choosing a coding system. The paper also details the relationship between population size, critical schema length, problem constraints and convergence.