A Machine Learning-Based Web Application for Student Performance Prediction Using Django Framework

Abstract

In recent years, the integration of machine learning techniques into educational systems has opened new possibilities for analyzing and predicting student performance. Early identification of students at risk of poor academic outcomes allows educators to intervene effectively and improve learning results. This project presents a web-based application developed using the Django framework that predicts student academic performance using multiple machine learning algorithms.The system utilizes a dataset containing various student attributes such as demographic details, family background, academic history, and behavioral factors. These features include age, parental occupation, study time, number of past failures, participation in extracurricular activities, internet access, and previous grades. Before training, the dataset undergoes preprocessing steps such as handling missing values, encoding categorical variables using Label Encoding, and feature scaling using StandardScaler. These steps ensure that the data is clean and suitable for machine learning models.The application implements several classification algorithms, including K-Nearest Neighbors (KNN), Random Forest, Support Vector Machine (SVM), Gradient Boosting, Logistic Regression, and XGBoost. Each model is trained and evaluated using standard performance metrics such as accuracy, precision, recall, and F1-score. The system compares these models and identifies the most effective one for predicting student outcomes.A user-friendly web interface allows administrators to upload datasets, train models, and visualize performance metrics through graphs. Users can input student details through a form, and the system predicts whether the student is likely to pass, fail, or drop out. The results are displayed along with warnings or suggestions to improve academic performance. Additionally, graphical representations such as pie charts and bar charts help in understanding overall trends and algorithm comparisons.The integration of machine learning with a web-based platform ensures accessibility and usability for nontechnical users such as teachers and academic administrators. The system is scalable and can be adapted to different educational datasets. By providing actionable insights, this application contributes to improving student success rates and reducing dropout rates.Overall, this project demonstrates the effectiveness of combining data science techniques with web technologies to build intelligent educational tools that support decision-making in academic environments.