An attempt to use a type-2 fuzzy logic system to assist with real estate appraisals

A Takagi-Sugeno-Kang type-1 fuzzy system to assist with real estate appraisals was developed with the aid of experts and comprised input variables relating to main attributes of a property being appraised. The rule bases of TSK zero-order fuzzy systems were generated using evolutionary algorithms. In order to reflect the uncertainty of available data, the type-1 fuzzy model was extended to the type-2 one using a singleton function as the second degree membership function. The comparative study of both models was carried out. The experiments were conducted using actual data.