Acadlore takes over the publication of IJEI from 2025 Vol. 8, No. 5. The preceding volumes were published under a CC BY 4.0 license by the previous owner, and displayed here as agreed between Acadlore and the previous owner. ✯ : This issue/volume is not published by Acadlore.
Open AccessMSW-DeepStack: Innovative Municipal Solid Waste Prediction Model for Informed Decision-Making
View Full Article|
Download PDFAbstract:
Municipal solid waste (MSW) is a fundamental problem in today’s urban environments, as its composition and quantity are constantly shifting due to many different influences. Sustainable waste management solutions could not be developed without reliable estimates of future waste generation. Predicting the amount of waste generated might assist authorities with decision-making and new technological approaches, such as machine learning and deep learning. In this study, a stacking ensemble of three models, namely, Grid Search Optimized XGBoost (GSO-XGBoost), Gated Recurrent Units (GRU), and Random Forest (RF) was proposed. The proposed MSW-DeepStack model outperforms the state-of-the-art algorithms by obtaining the highest R2 values ranging between 0.61 and 0.9. Furthermore, the MSW-DeepStack model obtained the lowest error rates: MAPE (0.1%-10.6%), MAE (0.0163-0.1182), RMSE (0.0014-0.1225), and ME (0.0022-0.213). The proposed MSW-DeepStack model has superior results compared to the state-of-the-art models, thereby demonstrating its efficiency and sturdiness. Further, the proposed model predicted that Singapore would generate around seven million metric tons of MSW by 2030. This estimation would aid in improving the MSW management methods and assist the authorities in making well-informed choices by shedding light on long-term trends in waste production.
