How to plot objective functions with number of generations in multiobjective genetic algorithm optimisation?

How to plot objective functions with number of generations in multiobjective genetic algorithm optimisation? Does number of generations are same as number of function count? Or number of generations are same as number of function values in Pareto front?

Answer 1

[Alan Weiss]

Each iteration of the solver is one generation. It involves making the new population. See How the Genetic Algorithm Works.

Each time the solver evaluates the fitness function counts as one function evaluation, the function count.

The number of individuals on the Pareto front is just that, and is not related to the other two things you mentioned.

To plot the various objective functions for just two functions you can use @gaplotpareto. For more objective functions, you probably have to write your own custom plot function.

Good luck,

Alan Weiss

MATLAB mathematical toolbox documentation

[RAVITA LAMBA]

Thank you sir for the reply. I couldn't get your point clearly. Suppose for my multiobjective optimisation, I have defined 400 generations and run the multiobjective GA optimisation tool. During running time, the function count keeps on increasing continuously and optimisation stopped after 250 generations. The stopping criteria is "Optimization terminated: average change in the spread of Pareto solutions less than options.TolFun". Now the total function count is suppose 5408. The number of iterations are same number of generations at which optimisation terminated which is 250. The Pareto front- function values and decision variables are 18 each. So I want to clear what is 250 and what is 18?

[Alan Weiss]

I have no idea what you are asking, sorry. You said "The number of iterations are same number of generations at which optimisation terminated which is 250." So why do you ask what is 250? You said "The Pareto front- function values and decision variables are 18 each." So why do you ask what is 18?

Alan Weiss

MATLAB mathematical toolbox documentation

[RAVITA LAMBA]

Thank you sir for your kind reply. I want to clear whether I am right or not what I have said above. My question is that I want to save the values of number of variables and functions at each iteration. So how to save these values at each iteration?

[Alan Weiss]

You asked the same question here, and I thought you understood my answer. For more help, look at the first answer here, or look at this answer.

Alan Weiss

MATLAB mathematical toolbox documentation

[RAVITA LAMBA]

Thank you sir for your kind reply. I want to plot this type of curve. But I do not have function values at each generation.

[Alan Weiss]

You can get the population at each generation. From that, you can calculate the fitness function or functions.

Alan Weiss

MATLAB mathematical toolbox documentation

[RAVITA LAMBA]

Thanks a lot sir for your help.