2 inputs data training for ANN

I am creating a neural network in MATLAB. My system has 2 inputs and the output is based on the 2 input values.

How do I define the inputs and feed them into the network?

The input 1 is array with 56 values and input 2 is also an is array with 56 values, how do I correctly define both inputs? And then how do I feed them into the network? The output value is also an array with 56 values.

Also, how many hidden layers are suitable for such model?

Answer 1

[Sam Chak]

@Krishna Saboo

For a shallow neural net, you can try something like this:

X = [1.0000    0.9019    0.6952    0.4712    0.2781    0.1341    0.0394   -0.0141   -0.0381   -0.0432   -0.0381   -0.0289   -0.0193   -0.0112   -0.0052   -0.0014    0.0008    0.0017    0.0019    0.0016    0.0012;... 
    0   -0.3438   -0.4529   -0.4273   -0.3396   -0.2368   -0.1445   -0.0736   -0.0257    0.0024    0.0158    0.0197    0.0181    0.0142    0.0097    0.0058    0.0029    0.0009   -0.0002   -0.0007   -0.0009];

T = [-1.0000   -0.4157   -0.0546    0.1331    0.2022    0.2009    0.1649    0.1183    0.0745    0.0398    0.0157    0.0011   -0.0063   -0.0089   -0.0086   -0.0069   -0.0049   -0.0030   -0.0016   -0.0006    0.0000];

net = feedforwardnet(10);
net = configure(net, X, T);
[net, tr] = train(net, X, T);