<![CDATA[The trend of refugee arrivals in Indonesia has become a pressing issue due to global conflicts, wars, and human rights violations. Accurate forecasting of these arrivals is crucial for effective planning and resource allocation by governmental and humanitarian organizations. This study aims to forecast refugee arrivals in Indonesia using exponential smoothing models, providing reliable predictions to support evidence-based decision-making and strategic policy planning. The dataset used in this research was obtained from the official UNHCR Indonesia repository, covering the period from April 2020 to July 2023. An exponential smoothing framework was employed, incorporating both single-parameter and two-parameter (Holt’s) models. The smoothing constants were optimized at α = 1.00 and β = 0.1237509, representing the level (Ft) and trend (Lt) components, respectively. A quantitative evaluation using key error metrics showed that both single-parameter and two-parameter (Holt’s) models captured data patterns accurately. The two-parameter model outperformed the single-parameter model (MAPE = 0.66, MAE = 84.57, RMSE = 132.63) and projected arrivals of 12.005, 11.913, and 11.821 for the next three periods. These results indicate that Holt’s model effectively represents the temporal dynamics of refugee inflows and provides a data-driven framework to support evidence-based refugee management and strategic policy planning in Indonesia.]]>
Copyrights © 2025
