Implementation of Machine Learning and Ensemble Learning Techniques to Proactively Predict Students' Academic Success in the Realm of Higher Education

Predicting and analyzing student performance is a huge boon for both students and teachers. A number of analytical methodologies have been utilized in recent studies to forecast students' success; the CGPA is the end product of these analyses, which have been conducted for both the internal assessment as well as for the external assessment. Few well known machine learning algorithms have been suggested. Ensemble Learning along with Decision Trees, Support Vector Machine and Random Forest -obtained perfect 100s for Precision Value, Recall Value, as well as in F1 Scores in the evaluation of performances. Naive Bayes has worse performance in comparison, with F1 Scores of 66, Recall Value of 65, and Precision of 68. When contrasted with decision tree models, Support Vector Machines, Multi-Layer Perceptron, Ensemble Enabled Learning, and Random Forest, Naive Bayes, AdaBoost, and K-Nearest Neighbors all perform worse in terms of accuracy. The Naive Bayes classifier models the input distributions across classes and is very good in text categorization because it is a generative method. Since it fails to identify critical features for class differentiation, it cannot be distinguished from discriminative classifiers.