In order to speed-up the process of optimization with GA, we can change some of the properties and/or operations in GA. The methods which can be implemented in this way, are called Micro Genetic Algorithms (micro-GA).
For example, in a particular micro-GA, we can consider the following changes on GA:
- consider small population, e.g., 5 chromosomes, with random re-initialization
- ignore mutation step
- consider a simple and relaxed convergence criteria, e.g., once 95% of the bits of the chromosomes are similar to the best chromosome
- keep the best chromosome of each generation for next generation (elitism) once the population converged
