Abstract:

The purchase of housing is a multi-dimensional problem that requires people to meet their housing, health, safety and various socio-cultural needs with a specific budget. The developments in the methods of artificial intelligence now recognize the possibility of solving many problems, such as buying housing in everyday life. Artificial nerve networks and deep learning methods are one of the methods used for this purpose. In this study, a smart system has been developed to record and predict the price of the housing by collecting housing data for sale on the internet. For the purpose of the system developed, the data for the sale of housing of the province of Istanbul have been collected. The data is divided in such a way that it will be trained and tested with the 10-layer cross-verification method. The data was modeled with 17 different algorithms according to 176 subjects. The most successful price for 852 houses in the Ataşehir district of Istanbul was obtained with the Rastgele Forest Algorithm. The study has developed a system in which researchers, companies and consumers operating in the housing market can predict the housing prices. Furthermore, subsequent studies aimed at giving researchers a new perspective on the use of different methods of artificial intelligence or housing properties.

Keywords: