The prevalence of traffic accidents in Blitar Regency is driven by multifaceted variables, including poor road infrastructure, mechanical failures, and low compliance among road users. To support preventive measures, this study evaluates time-series forecasting models to project future accident trends. It provides a comparative analysis between Single Exponential Smoothing (SES) and Single Moving Average (SMA) methods, utilizing 52 months of historical data from January 2022 to April 2026. Model performances are validated using MAD, MSE, and MAPE error metrics. The empirical findings indicate that the Single Moving Average configuration with parameter n = 3 outperforms the SES model by delivering the lowest error values. This optimal model projects 42 traffic accident cases for the May 2026 period. The findings of this study are intended to assist relevant stakeholders in formulating data-driven traffic safety policies and mitigation strategies.
Copyrights © 2026