A New Crossover Mechanism for Genetic Algorithm with Rank-based Selection Method

Crossover function is one of the important procedures of a Genetic Algorithm (GA). Genes of each chromosome are mated through crossover operator to produce new offspring that are evaluated if it qualifies to the next generation based on the derived fitness value. The study introduces a new crossover mechanism called Cross Average Crossover (CAX) with rank-based selection method that contributes to a promising result of variable minimization process. Further, the study presented a simulation of the generic GA having two-point crossover operator and the improved GA having CAX and compared their results based on the number of variables removed after the specified number of generations. A total of 1,409 student leaver's records from one of the universities in the Philippines were utilized for the simulation. It is evident that GA, having new crossover function, yields the most number of variables removed which is 90 percent that outperformed the generic GA having 25 percent of variables removed with the same number of generations.