Machine Learning-XGBoost Analysis of Subjective Well-being Among Chronic Hepatitis B Patients

Abstract

This study examined the relationships among objective social support, subjective social support, social participation, self-efficacy, and subjective well-being in chronic hepatitis B (CHB) patients using the XGBoost machine learning algorithm. Data were collected from 253 CHB patients. Using an optimal hyperparameter search, the XGBoost model achieved a classification accuracy of 98.04%. The results indicated that objective support, subjective support, self-efficacy, and social participation significantly predicted subjective well-being. XGBoost also highlighted self-efficacy as the most crucial predictive factor. These findings emphasize targeted psychological interventions to enhance self-efficacy and social support among CHB patients.