This work evaluates the predictive accuracy of three machine learning models—Linear Regression, Decision Trees, and Gradient Boosting—for forecasting nutritional intake. Utilizing a comprehensive dataset derived from the National Health and Nutrition Examination Survey (NHANES), we analyzed key input features such as age, weight, dietary habits, and physical activity levels to predict daily nutritional consumption. Our results demonstrated that Gradient Boosting significantly outperformed both Linear Regression and Decision Trees, achieving a Mean Absolute Error (MAE) of 2.6 and an R-squared (R²) value of 0.91. In contrast, Linear Regression and Decision Trees showed MAEs of 4.5 and 3.8, respectively. These findings underscore the effectiveness of Gradient Boosting in capturing complex dietary patterns and highlight its potential for application in personalized nutrition and healthcare interventions.