AI-Powered Trade Forecasting: A Data-Driven Approach to Saudi Arabia’s Non-Oil Exports

• DOI: 10.20944/preprints202503.0804.v1
• Publication Date: 2025-03-13T00:38:17Z
• Authors: Musab Aloudah; Mahdi Alajmi; Alaa Sagheer; Abdulelah Algosaibi; Badar Almarri; Eid Albalawi

This paper explores the application of artificial intelligence (AI) in forecasting Saudi Arabia’s non-oil export trajectories, aligning with Kingdom’s Vision 2030 objectives for economic diversification. A range machine learning models, including LSTM, Transformers, Ensemble Stacking, Random Forest, and XGBRegressor, were employed to analyse historical GDP data. Among these, Advanced Transformer model demonstrated superior predictive accuracy, achieving a MAPE 0.73%, underscoring its ability capture complex temporal dependencies. The non-linear Blending Ensemble, which integrates AdaBoost, exhibited robust performance (MAPE 1.23%), leveraging diverse strengths base regressors. Additionally, Temporal Fusion (TFT), incorporating data, provided insights into macroeconomic influences, though higher 5.48%, indicating challenges associated integrating indicators models. An analysis feature importance, utilizing SHAP Partial Dependence Plots, revealed that recent values (lag1, lag2, lag3, lag10) most influential predictors. These findings reinforce transformative role AI-driven methodologies, offering data-driven support decision makers strategists formulating policies promote sustainable growth.