Forecasting Housing Prices by Using Artificial Neural Networks

Artificial Intelligence techniques have penetrated all different areas of life since their first appearances and the use of intelligence systems has had a big potential for multi-disciplinary perspective by offering successful solutions to real world based problems. As there are many different techniques that can be examined under Artificial Intelligence research area, some of these techniques can become more efficient due to their wide solution scope. Artificial neural network (ANN) models are known remarkable solution ways among these techniques. The aim of this paper is to forecast housing prices by using ANN. Research data were collected from the 'sahibinden.com' web site, where houses are listed for sale advertisements in Turkey (particularly Usak for this research), and these data were used for training and testing period of the ANN model designed. As findings of this study, the ANN model estimated prices of houses for sale successfully, according to analyzes over the ANN model. Eventually, a decision support/adaptive suggestion mechanism is provided in this study, as an intelligent solution about optimum sales price of houses, which is needed by a large mass of society.