I want to access every individual for a generation before fitness function value is assigned so as to find the maximum of the fitness. In fact I need to find the maximum fitness for using it in a parameterless constraint handling operation where the worst feasible solution is assigned to the infeasible solution.
Depends on your dev environment. If you're using Matlab, you can simply create a 3D matrix for (NxMxG) [population_size, length_string_representation, num_generations]. For some other lower level languages you can use a struct.
Dear
This is a very simple task. you can save the value of fitness value during the evaluation of every individual and separate feasible from non feasible solutions. Then sort feasible solutions according to the fitness value the minimum fitness value can be penalized. More information about this issue can be found in my publication where I used similar concept and I am still using see :
.Mansour, N., Awad, M., and El-Fakih, K. (2006) Incremental Genetic Algorithm, International Arab Journal of Information Technology (IAJIT), vol. 3, no. 1, pp. 42-47.
- Badges
- Science topic