Assessing Student Quality of Life: Analysis of Key Influential Factors

The article focuses on the employment of a logistic regression model for feature selection, aiming to assess factors impacting student health and well-being. Recognizing the complexity of students’ well-being, our research employed a comprehensive questionnaire distributed among a cohort of 544 participants, featuring 201 carefully designed questions across seven thematic blocks. These blocks were tailored to explore various dimensions of students’ health and lifestyle, including physical health, mental well-being, academic stress, eating habits, etc. By leveraging machine learning techniques, the study meticulously selects the most relevant features from a dataset, analyzing their correlation with the target variable through F-value ANOVA. This process involves a systematic selection of top features, data transformation, and the division into training and testing sets, ensuring balanced representation of the target variable through stratified sampling. The logistic regression model is then trained and its predictive accuracy evaluated across varying feature sets, demonstrating the significance of feature selection on model effectiveness. The proposed method for feature selection is described and analyzed. The research highlights the model’s ability to identify key determinants of quality of life among students, emphasizing the role of healthy lifestyle choices on their overall well-being and academic performance. Apart from logistic regression, we conducted a comprehensive evaluation with 4 different classification models, and assessed their key metrics on predicting the well-being score.