Forecasting Indonesian Shrimp Exports for Business Strategy and Sustainable Blue Economy Policy: Evidence from Machine Learning

Fisheries are a strategic sector in Indonesia and are often linked to the country’s blue economy agenda. Shrimp remains a major export commodity, where performance is influenced by managerial and policy factors such as product quality compliance and cold-chain readiness, which are frequently discussed in relation to rejection risks in destination markets. This study provides a forecasting-based input for business strategy and policy by developing a machine-learning model to project Indonesia’s shrimp export trends and linking the results to blue economy policy analysis. XGBoost, CatBoost, and LightGBM were compared to identify the most suitable model. XGBoost produced the best results, with RMSE 1.87, MAE 0.48, and R² 1.00. In the first quarter, export values peaked in January, and whiteleg shrimp (udang vaname) dominated exports. The findings indicate that forecasting can support more targeted export planning, including aligning quality control and cold-chain capacity with peak periods, strengthening market coordination, and improving trade cooperation. Overall, this study highlights how predictive insights can inform practical strategies and policy direction while remaining aligned with sustainable blue economy goals.