Your brain is a neural network of neurons that receive inputs and decide whether or not to fire an output through a synapse. If a neuron fires, other neurons receive the output through the synapse and decide whether to fire themselves. However, the neurons in your brain are really much more complicated than this. For example, neursons have both proximal dendrites and distal dentrites that receive input signals Jeff Hawkins and Subutai Ahmad at Numenta found that the distal dendrites of a neuron condition it to recognize patterns of proximal dendrites. As a result, a neuron can recognize many different patterns.
Computer neural nets are made of data structures that approximate neurons. These artificial neurons have inputs and an output. The data structures include variables that specify what a neuron's output is based on its inputs. These variables are modified by algorithms that match the neural network's response to the training data.
The data structures also define how the inputs and outputs of the neurons are connected. Numerous processor cores set up inputs to the net, and calculate outputs. The neural network is organized with input layers, hidden layers, and output layers. A greatly simplified network is shown below.
