RESIDENTIAL PROPERTY PRICE FORECASTING USING FUZZY REGRESSION MODEL

Estimation of property sales prices has been addressed by several studies with an increasing interest in discovering the underlying non-linear relation that links the property features to the price of the property. Among the recent approaches applied were the advanced non-linear models such as Artificial Neural Network (ANN), Fuzzy Logic (FL) and Fuzzy Inference System (FIS), that showed practical success in modeling the nonlinear relation between various price predictors and the price value. Differently from the classical regression, Fuzzy Regression is a fuzzy logic-based regression model that has been effectively employed in multiple prediction problems. In this paper, we propose a new approach for sales price prediction using fuzzy regression model applying fuzzy sets and fuzzy logic techniques. The results clearly demonstrate an accurate prediction performance in comparison with artificial neural networks and reveal the implied fuzziness of the effect the independent variables have on the dependent variable.