VITAL SIGNS-BASED ENGINEERING STUDENTS’ STRESS LEVEL PREDICTION AND IMPACT ANALYSIS ON ACADEMIC PERFORMANCE USING A2RQNFIS

Abstract

For improving the engineering students’ academic performance, their stress level is predicted. But, the prevailing works didn’t focus on calculating engineering students’ stress index based on important vital signs. Thus, vital signs-based engineering students’ stress level prediction and impact analysis on academic performance using Adaptive Arctangent Rational Quadratic Neuro-Fuzzy Inference System (A2RQNFIS) are presented in this paper. Firstly, students' stress level data are taken; then, they are cleaned. Afterward, by using Density-Based Bray-curtisMatusita Spatial Dunn Hubert Clustering of Applications with Noise (DB2MSDHCAN), pattern analysis is performed. Thereafter, by employing the ErlangGompertz Kalman Filter (EGKF), temporal dynamic analysis is done. Similarly, from cleaned data, a spider plot is constructed. Next, from the plot and temporal dynamic analysis outcomes, features are extracted. Likewise, based on the vital signs, the stress index is calculated. Similarly, from Students’ Academic Stress Questionnaires (SASQ), the stress score is estimated. Then, the stress level of engineering students is predicted based on the extracted features, stress index, and stress score by using A2RQNFIS. Next, by using coping strategies, stress management suggestions are provided. Similarly, by using A2RQNFIS, the academic performance of engineering students is predicted. Afterward, by using the Pearson Correlation Coefficient (PCC), correlation analysis is done. Students are highly impacted if the correlation is high. Students are less impacted if the correlation is low. As per the results, the proposed model achieves a high accuracy of 98.68%, which is superior to the prevailing techniques.