Neural networks are automated structures, which are created by the example of natural sensory chains that form (set up) the animal brain. That sort of the structures learn to complete the assignments, by looking through the samples, without being directly programmed.
Neural Networks comprise a series of united elements or nodes, named artificial neurons that easily imitate themselves in an organic substance. Each contact can transfer one signal from one AN(artificial neuron) to another. The imitation neuron that gets the indicators may convert it and gives signs to neurons united with it.
The signal in the contact among neurons may be a real sum. The finished product is generated by a set of non-continuous functions of some original material. The layers are the basement of the artificial neurons.