Difference between revisions of "Examples"

(Prediction)
(Logical Functions)
Line 1: Line 1:
 
== Logical Functions ==
 
== Logical Functions ==
   
Training an artificial neural network to learn logical functions is just interesting in learning the use of such a network, but not for practical use.
+
Training an artificial neural network to learn logical functions is just interesting in learning the use of such a network, but not for practical use. The only interesting thing behind learning the XOR function is that in history of development of neural networks it was figured out the XOR function cannot be learned by just one neuron. But in the past it was quite difficult mathematically to find a solution to connect a few neurons together.
   
 
* [[logical XOR function]]
 
* [[logical XOR function]]
* [[logical OR function]]
+
* logical OR function
* [[logical AND function]]
+
* logical AND function
   
 
== Prediction ==
 
== Prediction ==

Revision as of 14:19, 21 December 2007

Logical Functions

Training an artificial neural network to learn logical functions is just interesting in learning the use of such a network, but not for practical use. The only interesting thing behind learning the XOR function is that in history of development of neural networks it was figured out the XOR function cannot be learned by just one neuron. But in the past it was quite difficult mathematically to find a solution to connect a few neurons together.

Prediction

One benefit of multilayer perceptron is the possibility of prediction.