Student Performance Prediction Using Artificial Neural Network

Abstract

Education and student both are correlated to each other. Education quality can be improved by student performance. If we want to lead a good life or productive life, then education is necessary and its quality needs to be improved. Performance evaluation of students is necessary for every educational institute in helping their student and teacher. In this study, we aim to predict the student category based on the performance of the student and propose a workflow of webbased four-tier architecture for the student performance prediction. For this purpose, a survey has been conducted on students in different universities in order to collect data and to analyze and predict the student category based on their performance. We proposed a new predictive model for predict student categories based on their performance and how a trained model can learn from realtime data to predict student performance. For categorized the student based on their performance, used multiple classification models using supervised machine learning algorithms. To get optimum features, we applied different data pre-processing techniques. Some supervised learning algorithm work well with all features yet. Each of the student category categorized by considering the top features. The analysis results indicate that we got the highest performance by using the Decision Tree Classifier (DT) by using op 10 features of Extra Tree Classifier Algorithm and XGBoost show the best performance with Chi-Square Feature Selection technique and other optimum selection features.