Design and Implementation of an Explainable AI Powered Model for Nutrition Assessment and Health Optimization

The increasing consumption of packaged foods hascontributed significantly to the rise of lifestyle-related health con-ditions such as obesity, diabetes, and hypertension, particularly inurban populations. Although nutritional information is mandatedon food labels, many consumers find these details difficult to interpret, limiting their ability to make informed dietarydecisions. This paper presents the design and implementationof an Explainable Artificial Intelligence (XAI)–based system fornutritional assessment and health optimization of packaged foodproducts. The proposed framework evaluates food items using key nutritional attributes, including sugar, sodium, saturatedfat, and calorie content, to compute a comprehensive healthscore. A linear regression model is employed due to its inherentinterpretability, enabling the system to explicitly explain the con-tribution of each nutrient to the final assessment. In addition tohealth scoring, the framework recommends healthier alternativeproducts by comparing nutritional profiles within a structuredfood database. By integrating transparency, interpretability, andactionable insights, the proposed approach enhances user trustand supports informed food choices, thereby contributing tohealthier lifestyles and aligning with Sustainable Development Goal 11.