I have some measurements as inputs which affect a single output and I want to use NN to study their relation and then predict the output as a function of the input. I found the neural fitting tool (nftool) but it is limited to a two-layer feed-forward network and 3 training algorithms (trainlm, trainbr, and trainscg).
In the help "Train and Apply Multilayer Shallow Neural Networks", I have also found 12 training algorithms with an example of a two layer feedforward network.
Is the nftool is limited to only 3 training algorithms and a 2 layers network? how can I use other algoritms in this tool ? I just want to make use of its benefits such as the plots of performance, error, regression, etc.