<![CDATA[Unemployment remains a persistent socioeconomic challenge in Indonesia, including Purbalingga Regency, Central Java. This study analyzes the unemployment trend and forecasts the number of unemployed individuals in Purbalingga Regency using a time-series approach. Annual unemployment data for 2010–2024 from the Central Bureau of Statistics (BPS) were modeled using Brown’s Double Exponential Smoothing (DES), which is suitable for non-seasonal series with a linear trend. The smoothing parameter (α) was examined from 0.1 to 0.9, and model performance was evaluated using MAD, MSE, and MAPE based on in-sample fitted errors over the 2010–2024 period. The results indicate a fluctuating but upward trend, particularly after the COVID-19 period. The best-performing parameter was α = 0.2, producing the lowest MAD and MAPE; under this evaluation setting, MAPE was below 1%, indicating low in-sample error. Using the selected model, unemployment in 2025 is forecast at approximately 31,795 people. These findings suggest that Brown’s DES can provide a practical baseline forecast to support evidence-based labor market policy and regional economic planning, while the results should be interpreted with caution, given the linear-trend and univariate assumptions.]]>
Copyrights © 2026
