I have a neural network that i want optimize number of hidden layers and neurons in every layer using an optimization algorithm like Imperialist competitive algorithm. As you know i should set limits of every variable for optimization. I can have maximum 2 layers. Besides this my system select best composition of features (among 21 features) based on cost function of my neural network (5-fold cross validation on test data) using that optimization algorithm. So my optimization algorithm has 21+2 variables. 21 for features and 2 for number of neurons in every layer. What is your idea and what do you recommend for boundary of number of neurons in every layer? I'm using patternnet in MATLAB. As you know we have difference composition of features and neural structure in calling cost function by optimization algorithm.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.