<![CDATA[Accurate evaluation of house valuation is crucial. Misestimation of house prices creates serious consequences for a variety of stakeholders. House prices can be modeled as a function of their constituent attributes. House prices are fuzzy due to negotiation and unpredictable market conditions. This study aims to develop rules to predict triangular fuzzy numbers of price for a given new house by implementing locally weighted KNN-based fuzzy regression, and to compare its performance with possibilistic fuzzy regression. The dataset used is the house valuation dataset. Data is examined using the modified Cheng and Lee k-nearest neighbor fuzzy regression and Tanaka’s possibilistic fuzzy regression. It is found that the modified Cheng and Lee k-nearest neighbor fuzzy regression outperforms possibilistic fuzzy regression in predicting the triangular fuzzy number of the house prices. The best performance is achieved when data is trained using the modified Cheng and Lee k-nearest neighbor fuzzy regression using: = 29 nearest neighbors, Minkowski distance with exponent parameter = 1.6 and an unequal weighting scheme with r = 1.]]>
Copyrights © 2026
