Logical Evaluation of Student Performance Analysis with Marks and Behavior Using Modified Deep Learning Strategy

It is essential that institutions use Educational Data Mining and Deep Learning to recognize academically poor students and assist them by establishing various recommendation systems to improve their effectiveness. By revealing valuable undetected trends in a student's data, these tools guide pupils in making decisions about their future. Ensemble Artificial Neural Networks and Light Gaussian Mixture Models are proposed for use in this study's prediction of students' academic achievement. The main aim of this research also set out to identify a common thread among the most popular ANN methods and techniques. Findings showed that ANN is always employed in tandem with data analysis and data mining approaches, permitting research to evaluate the efficacy of their findings in assessing its efficiency of evaluating academic success. When compared to existing machine-learning and deep-learning models, the suggested ensemble model performed better on a wide range of criteria. The suggested approach accomplished a remarkable precision that was higher than previously suggested models.. Additional research should look into a larger variety of student activities, a larger sample size of students, and a longer time span.