UTILIZING NUTRITIONAL AND LIFESTYLE DATA FOR PREDICTING STUDENT ACADEMIC PERFORMANCE: A MACHINE LEARNING APPROACH

Abstract

Nutrition and lifestyle factors have an enormous impact on students' academic performance. Nonetheless, there is a shortage of machine learning models to predict students' academic performance based on their nutrition and lifestyle. This paper intends to fill those gaps based on an extensive dataset of various attributes, underlining the capabilities of advanced machine learning models in uncovering the complex relationship between nutrition, lifestyle factors and student’s academic performance. A cross-sectional study was conducted in Kalar Technical College, Garmian Polytechnic University in Kurdistan region - Iraq, that involved 500 undergraduate students whose ages range from 18 to 22 years old; the dataset contains demographic characteristics, dietary intake, physical activity, and anthropometric measurements. Various Techniques, tools, and machine learning algorithms such as logistic regression and decision tree classifiers were employed using Python's Scikit-Learn library; finally, Pre-processing of the data was carried out to ensure its suitability for analysis. The machine learning model that the authors developed showed promising prediction results. While the logistic regression model had an accuracy of 70.4%, the decision tree model excelled with an accuracy of 98.55%. Furthermore, exercise, BMI, and dietary intake notably impacted students’ academic performance.