Forecast of Municipal Solid Waste Generation in Shah Alam Using Linear Trendline, Moving Average, Seasonal Naive, and Holt-Winters Additive Methods

Abstract

Shah Alam, a rapidly growing city in Selangor is experiencing a rise in municipal solid waste (MSW) generation, which poses challenges to sustainable urban management. This study analyses MSW generation patterns in Shah Alam for 2024 and 2025. This study forecast the MSW using four forecasting approaches, Linear Trendline, three-month Simple Moving Average (SMA), Seasonal Naive and the Holt-Winters Additive method, to forecast MSW generation in 2026. Monthly MSW data spanning 24 months (January 2024 to December 2025) from 56 collection zones in Shah Alam range from a minimum of 17,165 tons to a maximum of 22,328 tons. All four models were implemented in MATLAB. Forecast accuracy was assessed using Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The results show that among the seasonal models, the Holt-Winters Additive method achieved a MAPE of 2.79%, representing a 14.41% improvement over the Seasonal Naive baseline. The Holt-Winters method is therefore selected as the most suitable forecasting approach, with a total MSW generation of approximately 281,710 tons forecasted for 2026. These results provide actionable insights for local authorities in strengthening MSW management infrastructure and developing long-term reduction strategies.