A Hybrid Mental Health Education: Recursive Feature Selection and Intersubjective Group Interventions

This article proposes a hybrid of the two above model mentioned. Evidence-based feature selection and inter-subject relationship integration are beneficial for the training of college students in mental health. A total of 162 subjects received all three of the treatments: MHE (mental health education), CT (counseling techniques), and GT (group therapy). Measurement of psychological functioning was also included, namely the SCL-90 and the Group Therapy Factors Checklist. The key predictive factors were chosen by Recursive Feature Elimination (RFE) with Random Forest. In distinction, other group therapy process factors, e.g., Existential Awareness and Group Cohesion were the most likely to obligatorily appear. CT and GT promoted well-being, as results showed. And GT is the most optimal. This model allows for an on-demand, scaled telemental healthcare solution.