Assessing Students' Perceptions of Academic Quality in Higher Education Using Deep Belief Networks

The main objectives of this study are to evaluate a perceived service quality measuring scale and assess college student satisfaction. Administratively, this service quality research focuses on the registrar, schools, faculty, library, residence halls, rector, sports facility, and health centre. Descriptive and causal studies examine the scale's reliability and dimensionality. The suggested methodology includes preprocessing, feature extraction, and model training. Mining outcomes increase with data pretreatment, which fixes missing values and removes noise. Four meta-heuristic optimization methodologies yielded critical student happiness predictors. The model was trained using a DBN. Compared to ELM and ANN, the recommended model has 93.46% accuracy. Results reveal that preprocessing and tuning improved the model's prediction performance. As shown in this proposed, data preprocessing and feature extraction are essential for service quality assessments. The approach suggests a robust foundation for measuring and improving college student services compared to standard methods.