Abstract: In the past and present generation, the change in lifestyle of young people, less number of physical activities and eating fast food led to an alarming increase of diseases especially with college students. The unhealthy diet of students is responsible for unbalanced weight. The body mass index is used to classify a person's mass or weight as underweight, normal weight, overweight and obese based on tissue mass (bone, fat, and muscle) and height. Underweight leads to malnutrition, vitamin deficiencies, decrease in immune system function, growth issues and development issues. Major risk of being overweight leads to type-2 diabetes. Machine learning algorithms can solve many health problems. The project suggests the creation of an electronic college nurse that can perform the same functions as a human college nurse. Efficient Machine learning algorithms such as Naïve Bayes, Random Forest or Decision Tree used to build ML models to train the educational datasets and predicts student health problems. All these algorithms’ results are analysed and compared to find the best algorithm to predict student health risk. Many educational factors such as age, gender, food habits, family issues, pressure, academic results, height, weight etc.… are considered to predict student mental health problems. In our project work we build an application software with ML model that can predict the student health problems based on health factors and also recommends suitable diet plan. System aims at providing personalized, evidence-based diet recommendations for the students. Project also analyses trends in health metrics and enabling dynamic, context-aware advice for the students. Proposed system is a real time medical system useful for educational sector and students and built using Microsoft tools such as Visual Studio tool and SQL Server tool.

Keywords: Data science, Naïve Bayes, SVM, Random Forest, GUI, Student, Mental Issues, Machine learning, fertilizer.

| DOI: 10.17148/IARJSET.2025.12143