Implementing of artificial neural network to manufacturing sector

Future is always unknown. This makes people and company predict future. Companies try to forecast new period sales and prepare all its plans in related to new sales demand by demand forecasting. More correctly forecasted demad, more successfull company. Artificial neural network works as human brain. The brain takes new decisions, makes new predictions under learned information at sample events. At artificial neural network metod firstly, network educated as brain and got learned. Yet this learned network can make predictions itself. As more experienced person can predict more correctly in life; well educated network can predict more truly. In this thesis, at first section literature research made. On articles structure of artificial neural network investigated. At second section artificial neural network explained elaborately. In practice section demand forecasting done for a company manufacturing machine, its sales changing with trend behavior and seasonality by artificial neural network. MATLAB programme used for this. At designed artificial neural network machine model, months, sales on related month accepted as inputs and coming year same month sales predictions accepted as output. Different network constructions investigated for minimum error.