COMPARISION OF HEDONIC MODEL AND ARTIFICIAL NEURAL NETWORKS TO PREDICT HOUSING PRICES: AN EMPIRICAL STUDY IN THE TR52 REGION
KONUT FİYATLARININ TAHMİNİNDE HEDONİK MODEL İLE YAPAY SİNİR AĞLARININ KARŞILAŞTIRILMASI: TR 52 BÖLGESİNDE AMPRİK BİR ÇALIŞMA

Author : Emine Nihan CİCİ KARABOĞA -- Ayşe Elif YAZGAN - Nezahat KOÇYİĞİT - Yasemin TELLİ ÜÇLER
Number of pages : 465-478

Abstract

Housing that meets the need for housing, which is one of the important necessities of humanity for centuries, is one of the most basic elements of community life. As the transactions on this market do not take place collectively in a central market, real value can not be calculated for any dwelling since it does not have the continuity feature and it is difficult to determine how the price of dwelling changes over time. In this context, in the study, using the methods of Hedonic Model and Artificial Neural Network (ANN), it was attempted to be revealed the contribution of each variable, which is effective in identifying the sale price of housings that are present in the region TR52,to sale price, and estimation performances of two methods were compared. As a result of the application, in estimating housing prices, it was seen that ANN Model was a more effective model and a better alternative compared to Hedonic model.

Keywords

Housingmarketing, hedonic model, Artificialneural network

Read: 812

Download: 261