College Students' Mental Health Prediction Model Based on Time Series Analysis

Abstract

Social attention has long been focused on students' mental health, and forecasting mental health may be compared to a time-series classification challenge. The psychological health education programs at all of the colleges are now addressing the same problem, despite the fact that the number of schools is expanding quickly and the demand for students is always rising. In this Manuscript, College Students' Mental Health Prediction Model Based on Time Series Analysis (CS-MHP-CSTGCN) is proposed. Initially, the data is collected from Students Mental Health Assessments dataset. Then, the collected data is fed into pre-processing utilizing Implicit Bulk‑Surface Filtering (IBSF). The IBSF is used for data cleaning and data normalization. Then the preprocessed data are given to Continual Spatio-Temporal Graph Convolutional Networks (CSTGCN) for predicting a Mental Health of College Student’s and classify as normal and abnormal. CSTGCN does not express adapting optimization strategies to determine optimal parameters. Hence, the Multiplayer Battle Game-Inspired Optimizer (MBGIO) to optimize CSTGCN which accurately predict the Mental Health of College Student’s. The proposed CS-MHP-CSTGCN approach is implemented in Python. The suggested method's effectiveness was evaluated using performance measures such as MSE, F1-score, accuracy, precision, and recall. The suggested CS-MHP-CSTGCN approach contains 27.26%, 29.41% and 13.26% higher accuracy ,26.26%, 21.41% and 23.26% higher precision and 19.29%, 14.31% and 21.26% less mean squared error likened with current methods, like Student Behavior Prediction of Mental Health Based on Two-Stream Informer Network (SBP-MH-TSIN), Research on the Prediction and Intervention Model of Mental Health for Normal College Students Based on Machine Learning (PMH-NCS-RPMM) and Mental Disorder Prediction Model With the Ability of Deep Information Expression Using Convolution Neural Networks Technology (MDP-ADI-CNN) respectively.