A study on the factors influencing university students’ subjective well-being based on machine learning methods

Subjective well-being is a key area of research in psychology. Based on survey data from 384 university students, this study employed automated machine learning methods to construct a predictive model of subjective well-being, in which the Support Vector Machine (SVM) model performed best, achieving an overall prediction accuracy of 81.81%. The results indicate that the overall subjective well-being of the current university student population is relatively high; depression and interpersonal sensitivity are the most significant influencing factors, followed by hostility and paranoid ideation, among others. Family emotion also has a significant impact on subjective well-being. Based on these findings, fostering positive psychological traits and optimizing family functioning are suggested as approaches to enhance university students’ subjective well-being, thereby promoting their academic achievement and mental health development.