MATLAB: Editing the Individual during the FitnessFcn of a multiobj GA

fitnessfunctiongenetic algorithmGlobal Optimization Toolbox

Is there a possibillity to edit the current Individuals of a multiobj GA during the FitnessFcn? I ask, because the GA would converge much faster if I could adjust the current Individuals.

Best Answer

You should not change the individuals within the fitness function--this function is supposed to evaluate the fitness function, not change the individuals.

You are free to write custom mutation and crossover functions to generate new individuals however you like.

Good luck,

Alan Weiss

MATLAB mathematical toolbox documentation

Related Solutions

MATLAB: Hybrid Scheme in the Genetic Algorithm.

First one:

Undefined function or variable 'dejong2fcn'.

Second one: you have not properly passed the solver names in.

fminuncOptions = optimoptions(@fminunc,'Display','iter','Algorithm','quasi-newton');
options = optimoptions(@ga,'HybridFcn',{@fminunc, fminuncOptions});
options = optimoptions(options,'PlotFcns',{@gaplotbestf,@gaplotstopping});

MATLAB: Ga-fitness

variant

FitnessFcn = @(x)bsxfun(@minus,abs(bsxfun(@times,A,x)),B)

Best Answer

Related Solutions

MATLAB: Hybrid Scheme in the Genetic Algorithm.

MATLAB: Ga-fitness

Related Question