Before #8 , we need to prove that we're obtaining the best results possible. We need thus to check that the parametrization of the algorithm lets us obtain them. We need to check
Mutation operator: so far it's only random mutation, we should check that against mutation that changes values by +-1 (or whatever increment is appropriate for the values).
Crossover operator: there are two options, single and two-point crossover.
Population size: we should try 16, 32 and maybe 64
Termination condition should be taken out of the equation by simply stopping when an optimal solution is found, or a number of optimal solutions are found (1/4 of the population)
There are thus 2 (random + incremental mutation ) x 2 (single/2 point crossover) x 3 (16,32,64 population) = 12 experiments we need to repeat, 15 times each.
Before #8 , we need to prove that we're obtaining the best results possible. We need thus to check that the parametrization of the algorithm lets us obtain them. We need to check
Termination condition should be taken out of the equation by simply stopping when an optimal solution is found, or a number of optimal solutions are found (1/4 of the population)
There are thus 2 (random + incremental mutation ) x 2 (single/2 point crossover) x 3 (16,32,64 population) = 12 experiments we need to repeat, 15 times each.