Prediction of Rental Prices for Apartments in Brazil Using Regression Techniques

Abstract

The real estate industry is one of the most price-oriented industries and tends to fluctuate. The objective of the paper is predicting the rental price for a house. In this study, a predictive model based on the factors that influence the rental price has been constructed. The dataset has thirteen features. Regression techniques such as Gradient Boosting regressor, Ada Boosting regressor, K-nearest Neighbor regressor, Partial Least Square regressor, Random Forest regressor, Decision Tree regressor and Multilayer Perceptron regressor were applied. A predictive model is built using the regression techniques, and to pick the best performing model by performing a comparative analysis on their performance scores obtained. The expected outcome of the models is measured using performance metrics such as Mean Absolute Error (MAE), Mean Square Error (MSE), Root Mean Square Error (RMSE) and R-square score (R²) metric. This paper explains house rental price prediction model with the help of machine learning and the dataset used in our proposed model.