how to use genetic algorithm for EEG channel selection of motor imagery signal?

usage of genetic algorithm for selection of EEG channels such that it helps in identification of motor imagery EEG signal channel ( example C3, C4 ) like that.

[Geoff Hayes]

arunnagiri - do you have a fitness function for your GA? What are your inputs/variables for your function? What will you try to minimize or maximize?

[arunnagiri murrugesan]

i have no idea about fitness functions but my input will be matrix format of 160 X 1500 X 16 where 160 -no of trials , 16- no of channels , 1500- time points of the signal. i have 16 channels , need to minimize the channels such that localization of channels as which channels gives major contribution of motor imagery signal.

[Geoff Hayes]

You may want to read about MATLAB and genetic algorithms first to get an idea of what is required to solve a problem using GAs.

[arunnagiri murrugesan]

if i am using mean sqaure error as fitness function then how to proceed ? i need mimize the chanels.

[Geoff Hayes]

mean square error of what? what are the inputs to your fitness function?

[arunnagiri murrugesan]

mean square error of loss or mis classification accuracy. Another doubt is whether we can basic functions like Rosenbrock's function ?