Real Property Value Prediction Capability Using Fuzzy Logic and ANFIS

Abstract

Real property valuation is important as it gives an overall idea of the cost involved in the property beforehand. It enables to create transparency in the transaction of the real estate market. Prediction of real property value is beneficial and has multifold advantages. Traditionally property valuation is executed through the use of sales comparison method, rental method, profit method to name a few. When there is a nonlinear relationship between parameters affecting the price of the property and output, then it becomes difficult to predict the value by traditional methods. Therefore, to deal with the non-linearity of the dataset, modern techniques are resorted to. The use of different Artificial Intelligence (AI) techniques such as Artificial Neural Networks (ANN), genetic algorithm, fuzzy logic, Adaptive Neuro-Fuzzy Inference System (ANFIS) is increasing in the real property sector. In the present study, the prediction model is prepared for selected geographical sectors of Pune city, in India by using the Fuzzy Logic and ANFIS. It attempts to provide a comparison of their output based on the accuracy of the developed model and also based on the characteristics of the available dataset. The study endeavors to develop a value prediction capability with the use of soft computing.