Matlab - Neural Network Toolbox - Help Needed

I'm quite new to the NN Toolbox in Matlab and I would appreciate some help.

I'm trying to do a fluid dynamics simulation (Input - flow velocities, flow rates, diameters etc. - Output - pressure drop). So far I managed to get the basics out of it. I have a data set ot Input parameters and Targets which I used for training the network. I was wondering if I could use also classification data or some other type of parameters as input apart of the numerical input params.

For example...apart of typical velocities, diameters, flow rates and so on, my data is also classified, defining fluid types and other specifications for each timestamp.

Could I implement this information somehow or this isn't possible with NN?

Answer 1

[Greg Heath]

Any numerical information that is correlated with the output can be used as an input.

If you have any doubts, you could design a net using only that input.

However, if that input is also correlated with other inputs, it's true importance for improving performance will not be evident.

Only by using it with the other inputs can it's importance be deduced.

Hope this helps.

Thank you for formally accepting my answer

Greg