Forecasting daily LRT Ampang line ridership using SARIMA model

Accurate forecasting of public transport ridership is crucial for effective urban transit planning and operations. In particular, short-term prediction of LRT ridership in Kuala Lumpur plays a pivotal role in optimizing scheduling and resource allocation. The main objective of this study is to develop a reliable statistical model for forecasting the daily ridership of the LRT Ampang line. This study employs the Seasonal Autoregressive Integrated Moving Average (SARIMA) model based on Box-Jenkins methodology to analyze daily ridership data obtained from the Malaysia Official Open Data Portal. The dataset covers the period from January 2023 to May 2025. Model identification and estimation involved stationarity checks, ACF and PACF analysis, and parameter selection using AIC and BIC. The final model selected, SARIMA(1,1,2)(0,1,1)7, demonstrated strong forecasting performance with a MAPE of 9.78%. Residual analysis confirmed model adequacy through the ACF plot of residuals and the Ljung-Box test. This study concludes that SARIMA is a robust and interpretable statistical method for forecasting daily LRT ridership. It offers a structured statistical approach that contributes to improved decision-making in public transportation planning through short-term ridership forecasting.