The increase in diseases like obesity, diabetes, cardiovascular disorders, and metabolic syndromes has created an emergency need for personal and effective dietary management systems in this fast-changing world. Traditional dietary management systems are mostly based on a certain group of individuals, their lifestyle, and habits; failed at considering the diversity in living patterns, medical histories, regional food eating habits, and other important factors. Our research presents a solution by providing an intelligent diet planning system, which is created by studying structured food habits, daily routine, and medical histories. The dataset generated from the survey includes statistical characteristics such as age, gender, height, weight, smoking behavior, alcohol intake, sleep duration, exercise habits, dietary patterns, history of allergies, medicine intake, and history of lifestyle diseases. Models including probabilistic classification methods, margin-based classification techniques, tree-based learning models, and ensemble-based predictive frameworks are used in the approach. Among all these evaluated models, the decision tree classifier presents the highest predictive capability for the dataset used in this study, while probabilistic and ensemble-based methods also show strong performance. The outcome of this research portrays the possibility of intelligent data processing systems in creating tailored nutrition planning and health monitoring.
Moitra et al. (Fri,) studied this question.