<![CDATA[The rapid expansion of the digital economy has encouraged the integration of computational intelligence into economic efficiency analysis, including nutritional economics. Traditional nutritional assessment methods often rely on descriptive or linear approaches that overlook complex interactions among nutrients. This study proposes a Data-Driven Nutritional Efficiency Index (NEI), defined as Health Score divided by Calories + Sugar + Sodium + Saturated Fat + ε, to measure how effectively food products deliver health benefits relative to harmful components. The objective is to develop a predictive and optimized framework using machine learning and metaheuristic optimization. A structured dataset of 12 food items described by 12 nutritional variables is analyzed. Random Forest regression with 30 estimators and maximum depth of 5 models’ nonlinear relationships, while Bald Eagle Search (population size = 6; iterations = 8) performs feature selection to reduce redundancy and prevent overfitting. Model performance is evaluated using Leave-One-Out Cross-Validation, RMSE, and R². Results show strong predictive agreement, with NEI values ranging from 0.114676 to 0.672911. Findings confirm that nutritional efficiency varies substantially and cannot be explained by calorie content alone, supporting scalable, data-driven, AI-enhanced nutrition decision systems.]]>
Copyrights © 2026
