I have a feedforward network architecture like this= net=newff(​minmax(inp​ut),[20 10 2],{'tansig', 'logsig' , 'tansig'},​'trainlm',​'learngd',​'mse'); Can any body explain this in detail?? remember that i have 3 output classes e.g class A,B,C.

I have a feedforward network architecture like this:

   net=newff(minmax(input),[20 10 2],{'tansig', 'logsig' , 'tansig'},'trainlm','learngd','mse');

Can any body explain this in detail?? remember that i have 19 input instences and 3 output classes e.g class A,B,C.

Is it mean that there are 20 input neurons? 10 hidden neurons? and 2 output neurons? need explanation?

Basically i want to produce three output and my system produces it correctly but the key to understanding is not know that if the out put neurons are two then how it produce 3 outputs??

The heading of ( Layer Layer Layer ) on the following picture shows how many input, hidden and output layers??

And at the end of picture what does output 2 mean?? while the predecessor 2 in last layer already shows the output neurons.