C="Neural Net Xor example for the Back Propagation method.
To start training the net, select Simulate/Go from the menu above. When the error in the lower plot block goes to zero, stop training by selecting Simulate/Stop. Click the right mouse button over the xor.net block, and turn off \"Learn\", Click on OK, and rerun the simulation. You will see the learned xor results come out of xor.net, and see the difference between the neural net results and the built-in xor operator results in VisSim in the upper plot.
If you select the BP/Momentum learning method, you will see that the net learns significantly faster.
The Back Propagation methods can be trained for long periods of time. 10 to 24 hour periods are not uncommon for large, complex data sets.
The xor function is a simple, but classical example of a neural net learning a nonlinear function. At one time, it was thought that neural nets were not capable of learning the simple xor function."