Machine learning prediction of sleep quality among UiTM Perlis students using the Pittsburgh sleep quality index

Sleep serves as a fundamental process for repairing the body, replenishing mental energy, and maintaining emotional stability (Zhang et al., 2023). However, a low quality of sleep is typical of the university students. This study aimed to assess sleep quality among UiTM Perlis students using machine learning (ML) and to compare the performance of XGBoost, Random Forest, and Decision Tree models in predicting sleep quality. A total of 242 responses were collected using the Pittsburgh Sleep Quality Index (PSQI) questionnaire (Buysse et al., 1989), which included demographic and lifestyle factors. After encoding, scaling, and preprocessing, all models were tuned using GridSearchCV with 5-fold cross-validation. Performance was evaluated using accuracy, precision, recall, F1-score, and AUC. The mean PSQI global score was 7.18, indicating that most students had poor sleep quality. Among the models tested, XGBoost achieved the highest test accuracy (97.96%) and best cross-validation score (95.34%) with the lowest standard deviation, confirming its accuracy and stability. Overall, this study demonstrates that ML models, particularly XGBoost, can effectively predict sleep quality among UiTM Perlis students.