Predicting Student Well-Being Through Academic Performance and Daily Habits

Abstract: Over the past several years,there,has been an increase in the number of cases where students report higher levels of mental health strain due to various reasons such as pressure in studies,changes in lifestyle,and economical issues among others.It is necessary to find out who the students are that might be depressed and give them appropriate treatment.In this study,an attempt was made to develop a predictive model for the probability of depression in students as per the academic and lifestyle indicators of the Kaggle Student Depression Data Set.The Kaggle data set was pre-processed with missing value treatment,encoding of labels and normalization of features.Broadly,machine learning is categorized into four types namely Logistic Regression,SVM,Random Forest and XGBoost were developed using the Kaggle Student Depression Data Set.The metrics analyzed were accuracy,precision,recall,and F1-score.The accuracy of logistic regression model was the highest (83.61%) among four machine learning models.

Keywords: Machine Learning,Student Depression Prediction,Student Mental Health,Academic Factors, Lifestyle Fac- tors,Logistic Regression,Predictive Modeling,Early Risk Detection.