i-Saturate: The New Discovery of Stopping Criterion in Genetic Algorithm

A stopping criterion for evolutionary algorithms like Genetic Algorithm (GA) is crucial in determining the optimum solution. It is common for a stopping criterion like maximum
generations or fittest chromosome repetition used in GA to solve hard optimization problems. However, these stopping criteria require human intervention to make certain changes. In this study, a new stopping criterion called i-Saturate that measures saturation of population fitness of every generation chromosome (in GA searching process) is reported. The searching process would stop when the fitness deviation of the population was small. A model using fittest chromosome repetition was developed to compare the efficiency with i-Saturate. It was found that the performance of the developed model was good at the low mutation rate (0.01,0.02) but
the i-Saturate model was better when mutation rate was greater than 0.03. The probabilities of the i-Saturate model finding global optimum solution were very close to 1 when mutation rate was above 0.07. It was concluded that the i-Saturate model has demonstrated better searching ability than the comparative model and it intelligently stops searching without human intervention.