Current Proceedings on Technology
Yazarlar: Ozcan Asilkan, Adriola Faqolli, Alba Gerdeci, Betim Cico
Konular:-
Anahtar Kelimeler: House Price Estimation,Hedonic Regression,Artificial Neural Networks,Tirana
Özet: Over the last two decades, a large number of Albanian inhabitants have migrated toward the capital, Tirana. Due to this phenomenon, the demand for housing in Tirana has increased significantly. Addressing to this problem, this paper examines the applicability of Hedonic Regression (HR) and Artificial Neural Network (ANN) models in housing market in Tirana. According to HR model, estimation of the implicit prices of the housing attributes determines the price of the complete house. Typically, more of the desired attributes the house contains, greater its price is. ANN model produces an estimated house price based on a particular input set. Then by comparing the model output to the actual output, the accuracy of this value is improved. In this paper, 200 properties in Tirana have been sampled to illustrate the applicability of these models. The results show that most of the coefficients are highly significant, thus supports the potential of both models on house price estimation. ANN can even be a better alternative to HR to estimate the house prices in Tirana.