Addis Ababa City House Price Prediction By Using Regression Algorithm

Abstract

House is a basic need for people and used to measure the economy of individuals, organizations, and country. However, the unavailability of the exact sales price of houses in Addis Ababa city is not well known. The aim of this study is to predict the sales price of the house by using a regression algorithm for Addis Ababa city house seller and house buyer with respect to their budgets and priorities by analyzing previous market trends and house price ranges. In order to choose a prediction method, several regression models are explored and compared to find the best performance solution for house price prediction. Methods that has been discussed in this studyinclude ridge regression, elastic regression, k-nearest-neighbors(KNN), Neural network(NN), random forest, and decision tree. All this regression algorithm technique was compared with results of mean absolute error(MAE) and mean absolute percentage error (MAPE) to select the type of regression algorithm with the least MAE and MAPE. Result depicts that from all other algorithm techniques decision tree algorithm show minimum error of MAPE of 9.81 % while ridge algorithm technique resulted with maximum error of 17.50 %. This show that decision tree algorithm techniques best fitted house sales price for Addis Ababa city with minimum error of 15.76%.