The Comparison Study of Regression Models (Multiple Linear Regression, Ridge, Lasso, Random Forest, and Polynomial Regression) for House Price Prediction in West Nusa Tenggara

Predicting house prices is essential in the real estate industry as it enables stakeholders to make informed decisions. Accurately predicting house prices is paramount in facilitating buying and selling transactions, aiding property valuation, and providing valuable insights for investors and homeowners. This research compares five regression models, namely Multiple Linear Regression, Ridge, Lasso, Random Forest, and Polynomial Regression, to predict house prices in West Nusa Tenggara Province. These considered building area, land area, number of bedrooms, and bathrooms. The data was taken from the public website, namely Lamudi Website, and was collected using a web scraping method and then processed using Machine Learning. By considering factors such as building area, land area, number of bedrooms, and bathrooms, the study aimed to determine which model delivered the most accurate predictions and exhibited a low error rate. This research used various metrics such as R-Squared, Root Mean Square Error (RMSE), and Cross-Validation to measure the accuracy of these models. In this study, the R-Squared and RMSE methods showed that Multiple Linear Regression and Lasso Regression were the best models with the same R-Squared and RMSE values, namely R-Squared = 0.6947 (69.47%) and RMSE = 2863760831. Meanwhile, the RMSE (Cross-Validation) method shows that Random Forest Regression is the best model with a value of 3343572297. These research outcomes have the potential to provide valuable guidance to the local community in making informed decisions regarding property transactions in West Nusa Tenggara.