The image above shows the ANN for solving XOR (bias input is not included).

These are confusing terms:

  • Input layer: Should be called input nodes as they are not neurons
  • First hidden layer: Should be called just ‘first layer’ as this is the first neuron layer
  • Output layer: Should be called ‘last layer’ as this term is confusing with the output processor (eg. round up value) after it.
  • Output processors are some kind of classes or functions to form up the final results.