Student Academic Performance Detection using Gaussian Distributive Genetic Feature Optimization Based Congruence Gradient Deep Multilayer Perceptron Classifier

Abstract

Student educational result prediction is fundamental aspect for detecting hidden association in educational information and forecasting students' academic attainments. The main aim of applying ML for student academic performance forecast is to enhance the performance outcomes by providing timely intervention. Several machine learning techniques are employed but it faced challenges in accurate prediction.  The Gaussian distributive genetic optimization based Congruence gradient Deep Multilayer Perceptron (GDGO-CGDMP) model is proposed to improve accuracy of student academic performance forecast. It comprises of data acquisition, data pre-processing, feature extraction, classification. First, number of student data samples is collected from the dataset during the acquisition phase. In data pre-processing, split two stages namely missing data handling and outlier data removal. Jaccard indexive Elitist genetic algorithm is used to extract important features as of pre-processed database, thereby reducing time required for performance prediction. Classification using Congruence Correlative gradient optimized deep multilayer perceptive classifier to determine similarity or relationship between features, enabling the identification of performance level in terms of pass or fail and it enhances the ability to academic performance level and improved classification outcomes. Finally, student performance predictions are achieved with minimal error by applying adaptive gradient method. Experiments conducted on a 40,000-sample student dataset using GDGO-CGDMP achieves 95.11% accuracy, 95.33% precision, 97% recall, and a 96.15% F1-score, 3% and 6% in accuracy compared to conventional methods, while reducing prediction time by 19%. The proposed framework described that superior scalability and robustness across varying dataset sizes, making it suitable for real-time deployment in educational institutions.