how to control and access which variables are considerent at each iteration in ga optimization?

Hello,

i am running a mixed integer general algorithm optimization problem with about 90 variables. I would like to control which of those variables are considered in each iteration. This is because I would like emphasize this small set of variables in the cost function such that their effect is more clear. The population and generation seems to be not what i am looking for. Anyone has any ideas? Mutation? Elite children? Much appreciated!

cheers, don

[mb]

as described here: http://www.mathworks.es/matlabcentral/answers/94510-how-can-i-obtain-and-view-the-automatic-generated-initial-population-from-ga-from-optimization-toolb

does this function provide the population list before or after the current iteration? How could I call this function from the fitness function?

Answer 1

[José-Luis]

I might be mistaken, but that's not how I would go about doing sensitivity analysis. It sounds to me that all you'd be achieving is crippling your optimization algorithm if you start messing with its internal state, unless you're quite certain about what you are doing.

[mb]

Hello José-Luis

and thank you for the answer. It would be sufficient for me just to know what variables are changed before the next iteration. I dont really know how the population list works in every iteration. I would be more interested to use this information in the fitness function...