Optimize the weights of neural network using genetic algorithm

how to optimize the weights in neural network using genetic algorithm.... I saw some procedure in mathworks but have a small doubt in it. regarding MSE error calculation.. net = setx(net, x'); % To evaluate the ouputs based on the given % weights and biases vector y = net(inputs)

% Calculating the mean squared error mse_calc = sum((y-targets).^2)/length(y); I want an explanation for y=net(inputs) what does it retrieve from the above operations? for me it retrieves [1*1]vector. how can i subtract it from targets which i have set it to .... % inputs for the neural net inputs = (1:9); % targets for the neural net targets = cos(inputs.^2)

% number of neurons n = 2; If anyone interested pls... answer me!!!!