Stress Prediction in Higher Education Students Using Psychometric Assessments and AOA-CNN-XGBoost Models

Research backs up the common perception that the years preceding up to college are the most difficult, showing that students report higher levels of discomfort and worse self-esteem during this switch. Positive personality qualities such as optimism, hope, and happiness were examined in this system along with 215 first-semester Israeli university students' ratings of functional impairment, psychological distress, and self-esteem. The three stages that make up this suggested method are data preprocessing, model training, and feature selection. In order to process data that would be infeasible to process without data preprocessing, the data can be adjusted to meet the specifications of each data mining technique. Beginning with different forms of feature classification or aggregation and progressing to individual activity levels, several levels of feature granularity were investigated in order to employ a feature selection technique. The proposed approach used an AOA-CNN-XGBoost for the whole model training phase. This novel method outperforms CNN and AOA with an average accuracy of 89.27%.