Predicting house sale price using fuzzy logic, Artificial Neural Network and K-Nearest Neighbor

Abstract

Determining the value of land and home are regularly determined at the earliest by the seller, however determining the right price in the sales process will affect the buyer's desire to elect and bid. Special characteristics in Indonesia, tax object value (NJOP) and location parameters are high influence to the price. In this paper we proposed the prediction of land and house value using several methods. Fuzzy logic, Artificial Neural Network and K-Nearest Neighbor are compared in this paper to discover the most appropriate method that can be used as a reference for determining the price by the sellers. Google Maps is used to represent the spatial data for prediction parameter. The variables that used in the methods are NJOP of land, the locations, the age, NJOP of house, and the valuable location of the land. The experimental methods are tested by comparing between the real price transaction and the prediction using MAPE formula.