MATLAB: What is the difference between nftool & nntool ?

nntool & nftool

Hello, All

what is the difference between nftool & nntool ?? i'm building a network that will be used for prediction….which tool is more accurate nftool or nntool???

Best Answer

nftool specifically uses a 2 layer LM for fitting.

nntool allows the creation of more general neural networks.

As the networks created by nftool are a subset of the networks that can be created with nntool, nntool properly applied will never be less accurate. But nftool might be as accurate as you need.

Related Solutions

MATLAB: How to train ann with multiple input data

% Create a network (i have used feed-forward backpropagation)
net = newff(TrainFeatures, Trainlabels, num_of_hidden_Neurons);
% Set Train, Test and Validation Ration
net.divideParam.trainRatio = 70/100;  
net.divideParam.valRatio = 15/100;  
net.divideParam.testRatio = 15/100; 
net = train(net, TrainFeatures, Trainlabels);

MATLAB: Nftool output transfer function

Hi Matthew,

The output of the neural network depends on the transfer function of the output layer. So, yes changing the output transfer function is a good idea.

The nftool is generally used for training a simple two-layer feed-forward network. You cannot change the output transfer function or load a custom made network in the GUI.

If you want to handcraft a regression network according to your requirements, you can look into https://in.mathworks.com/help/deeplearning/ug/train-a-convolutional-neural-network-for-regression.html which should give you more freedom for experimentation.

You can change easily change the transfer function using -

net.Layers{end}.transferFcn='logsig'

Hope this helps.

Best Answer

Related Solutions

MATLAB: How to train ann with multiple input data

MATLAB: Nftool output transfer function

Related Question