Abstract: A genetic algorithm is used to generate a reduced-order surrogate model (ROSM) that approximates a data set. A fitness value is determined, for each parent chromosome of a population of parent chromosomes, by applying the data set to a chromosome function constructed from a sequence of genes of said each parent chromosome. Each parent chromosome is then randomly assigned into either a first subpopulation or a second subpopulation. The first subpopulation is ordered based on their fitness values. A population of child chromosomes is then generated from the first and second subpopulations. The ROSM is represented by a chromosome function of the child chromosome having a highest fitness value of the population of child chromosomes. The chromosome function is constructed by parsing the genes of the corresponding chromosome, each of which stores a mathematical primitive. The ROSM may be outputted in human-readable format.