Controlling Crossover Probability in Case of a Genetic Algorithm

Genetic Algorithms (GAs) are commonly used today worldwide. Various observations have been theorized about genetic algorithms regarding the mutation probability and the population size. Basically these are the search heuristics that mimic the process of natural evolution. This heuristic is routinely used to generate useful solutions for optimization and search problems. GAs belong to the larger class of evolutionary algorithms (EAs), which generate solutions to maximize problem solving by using techniques inspired by natural evolution, such as inheritance, mutation, selection, and crossover. In this paper we study of a simple heuristic in order to control the crossover probability of a GA. We will also explain how stress factors in on the crossover probability and why it is an important phenomenon in case of a GA and how it can be controlled effectively. Experimental results show that, for reaching lower probability from higher probability, we can get faster optimal solutions for any problem. These experimental values are derived by taking the values at the high probability and then slowly yet steadily decreasing them.