Predicting the Quantity of Municipal Solid Waste using XGBoost Model

The quantity of Municipal Solid Waste (MSW) gets intensified, based on various factors such as population growth, monetary status and consumption patterns. The insufficiency of elementary trash data is a critical problem for managing the MSW. In this study, the goal is to forecast the MSW generation of Northern Ireland. A precise model was developed to estimate the total amount of waste produced for every quarter of the year, by employing the Machine Learning techniques. The seasonal ARIMA (s-ARIMA) and Extreme Gradient Boosting (XGBoost) models were employed to estimate the amount of waste produced. On comparing both the models, XGBoost performed better. Thus, the parameters of the XGBoost were tuned to yield the optimal outcome. The XGBoost with the tuned hyperparameters achieved an optimum result with the higher coefficient of determination (R2) value as 0.5325 and lower RMSE value of 13215.97. The prediction of the MSW weight would help the decision-makers in treating and disposing solid waste appropriately.