Traditional methods for municipal solid waste (MSW) forecasting rely mainly on simplistic models that fail to capture complex urban dynamics, particularly during disruptive events like pandemics. Current machine learning approaches do not include policy-driven variables and real-time adaptability for municipal automation systems. How can machine learning models be enhanced with interrupted-time-series analysis to quantify pandemic impacts and provide automation-ready MSW forecasting for smart city governance? In order to find the solution, this study compiled 199 months of MSW data (2009-2024) from Surat Municipal Corporation (SMC). A rigorous preprocessing is employed by engineering temporal lag features (1-12 months) and creating a COVID-19 binary indicator (March 2020-March 2022). In our study, six supervised regression models were evaluated: Linear Regression, Random Forest, XGBoost, Support Vector Regression, Stacking Regressor, and Multilayer Perceptron, using rigorous interrupted-time-series comparison with and without pandemic indicators. Random Forest achieved highest accuracy (R 2 = 0. 85), with lag₁ and residential waste as strongest predictors. The COVID-19 dummy reduced RMSE by 16% in Linear Regression and 10% in XGBoost thus demonstrating measurable pandemic impact quantification. SHAP analysis confirmed temporal dependencies and residential consumption patterns as the primary drivers. Our study systematically integrates interrupted-time-series methodology with the machine learning modeling for MSW forecasting which provides explicit pandemic impact measurement and automation-ready frameworks for municipal decision-making for their operations. Results we obtained demonstrate that hybrid ML + policy variable approaches enable resilient and interpretable forecasting systems capable of adapting to urban disruptions while supporting data-driven smart city waste management automation.
Rathod et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: