Artificial neural network is a mathematical model of human brain, which can be trained to perform a specific task based on available knowledge base. Its design loosely follows neural structure of a human brain. Human brain excels the conventional computers in three broad areas: pattern recognition, associative recalling, and learning.
Artificial Neural network is composed of three layers:
• one or many hidden layers
• output layer
Hidden and output layers include the combination of weights, biases, and transfer functions. The transfer functions are linear or non-linear algebraic functions. When an input is presented to the network, then weights and biases are adjusted so that a desired output is obtained. Once a neural network is trained to a satisfactory level, it can be used on generic data.