Estimation of sale of housing in Antalya with multiple linear regression models and artificial neural networks

Demand is, in general, the amount consumers are prepared to buy at a certain price level of a good or service. Demand forecasts are estimates of future periods of goods or services produced by businesses or individuals and affect future decisions. The artificial neural network frequently used to make predictions has been inspired by the working principle of the human brain and has an important place in artificial intelligence studies. Along with many predictive analysis, artificial neural networks provide quite good results in solving modeled and predicted power problems. In this study, it is aimed to estimate the demand for housing in Antalya with artificial neural networks. Monthly data between the years 01.2013-09.2017 have been used. For the analysis, 85% of the data is divided into training and the remaining 15% is divided into two parts as test data. First, it is tested by assuming the assumption with the multiple linear regression model of classical methods. Forecast obtained by artificial neural networks were than compared with the results of the regression analysis. As a result, artificial neural networks have been found to be more successful and effective in predictive analysis.