Analysis of Waste Transportation Volume in Jakarta Province using Linear Regression and Random Forest Regression

Abstract

The accumulation of waste volume in the river waters of DKI Jakarta is still a significant problem that cannot be solved optimally because the population continues to increase every year, so the tonnage of waste also increases as well as some residents of DKI Jakarta still throw garbage into the river. In predicting the level of waste volume, the DKI Jakarta Provincial Environment Agency must make decisions, so it is necessary to carry out a prediction stage regarding the increase in waste in the future. For this reason, this research performs a prediction stage by utilizing two machine learning algorithms: Linear Regression and Random Forest Regression. The experiment used historical data on waste volume transportation from January to June 2021. The experimental results showed that the Random Forest Regression had the lowest error values of 0.82 and 0.81, with a training and testing data ratio of 80%:20%. On the other hand, Linear Regression has an error value of 0.83 and 0.82 at a ratio of 80%:20%. The analysis discussed in this study can be a reference for predicting and taking the necessary actions to prevent an increase in the volume of waste in DKI Jakarta Province.