Prediction of Surabaya City Apartment Rental Price Using the Ensemble Method
Keywords:Rent Prices, Apartments, Bagging, Boosting, Stacking, Ensemble Learning
Increasing the need for residential space requires in city to improve development infrastructure to meet the needs of its inhabitants. Having an apartment will help optimize the capacity of residential land in limited land, especially for areas with high population density and small areas such as the city of Surabaya. Until 2020, the level of demand for housing or rental in apartments tends to decline and after the COVID-19 pandemic made a huge impact on the apartment market from a sales perspective and prices dropped dramatically. A property consulting firm, Coldwell Banker Commercial, stated that among other big cities, only Surabaya had a better performance in apartment marketing, because the sales rate and average price increased by 0.4%. Seeing the urgency and potential for the weakening of modern society in choosing apartment housing in the city of Surabaya, the authors conducted a study that could assist in predicting apartment rental prices according to apartment criteria in Surabaya such as sub-district location, unit type, area, floor number, and unit facility contents. The data used is data from travelio.com which is one of the most trusted websites in Indonesia for apartment rentals using the scraping method. The apartment rental price prediction model using the ensemble method found that the stacking model with a combination of gradient boosting regression, random forest, decision tree regression, linear regression meta liner regression has higher performance, with an MAE value of IDR 243,401, RMSE of IDR 179,432 and a value R2 is 63.28%, which means that it affects the price of apartment rentals by 63.28%, while 36.72% is influenced by other factors that are not included in the model.
Copyright (c) 2023 Erlina Komaruljannah, Chastine Fatichah
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.