PREDICTIVE MATHEMATICAL MODELS COMBINED WITH AI FOR DIGITAL TRANSFORMATION IN SUPPLY CHAIN MANAGEMENT

Supply chain systems are the lifeline of economies, and in Iraq between 2020 and 2024 their resilience depended on predictive mathematical models combined with AI. This study analyzed how forecasting techniques, optimization methods, and resilience modeling shaped transformation outcomes such as forecast accuracy, inventory reduction, delivery speed, and risk reduction under fragile conditions. A descriptive design was applied, drawing on secondary data from 25 sector-year observations covering logistics, retail, manufacturing, and energy. Correlation analysis revealed strong positive links between outcomes and resilience modeling at 0.80, forecasting at 0.77, and optimization at 0.73, while structural constraints had a negative effect at −0.59. Regression confirmed resilience modeling as the strongest driver with a coefficient of 0.36, followed by forecasting at 0.27 and optimization at 0.23, while structural constraints reduced outcomes at −0.20. The model explained 81 percent of the variance, proving the robustness of the framework. Results showed forecast accuracy improved from 55 to 78 percent, inventory waste reduced by 30 percent, delivery reliability rose by 20 percent, and resilience scores doubled from 30 to 70. The findings imply that predictive AI can transform fragile supply chains into adaptive systems, but scaling requires stronger infrastructure, better data governance, and institutional capacity. Recommendations emphasize investment in hybrid forecasting, reinforcement learning, and resilience modeling, alongside parallel reforms in digital infrastructure, workforce skills, and governance frameworks.