Explanation of neural network creating and training in MATLAB

Question

I am new to MATLAB and having a problem with the MATLAB code for creating and training a neural network. As an example,

x = [0 1 2 3 4 5 6 7 8];
t = [0 0.84 0.91 0.14 -0.77 -0.96 -0.28 0.66 0.99];
net = newff(x',t',5,{'tansig', 'purelin'});
net = train(net,x',t');

What I need to know is the relationship between two vectors x and t. I know x is input vector and t is target vector. But what is the relationship between the set of x values and t values. Does the neural network have 9 input nodes to take the values in the vector x as input values? Should the first value of the t vector (in this case, 0) be the output of the network after training the network only once? should the second value be considered as the output value after the second iteration of the training? Can someone explain me this. I am confused.

damienfrancois damienfrancois · Accepted Answer · 2014-05-19T20:05:00

The neural net has one input node. The first value of t is what the output of the (fully trained) neural net should be when fed with input 0, while 0.84 is the output it should give when the input is 1, etc.

Explanation of neural network creating and training in MATLAB

1 Answers