Abstract:

The demand forecast is one of the important activities that and individuals need to do in order to take strategies and measures toward the future. This activity is a process of correct and incorrect calculation of the future demand for a service or product, and today, when the competition is intense, it has vital importance in the production and service sectors for the sustainability of the products or services they provide. In addition to statistical prediction methods, artificial intelligence techniques are also effectively used today for demand prediction. In this study, the forecast of housing requests in the province of Antalya was obtained with the very variable linear regression analysis with the Eviews program. For the same set of data, the advanced feed-backed artificial nerve networks with the help of the Matlab program were also predicted and the validity of the results and the performance of the network were compared with the results achieved with the regression model. According to the comparison, the regression analysis has an average error value of 9%, while the artificial nervous network estimate model has an error value of 1%. Therefore, it has been found that artificial nerve networks are more successful for the regional housing sales model.

Keywords: