About the Project This project aims to predict student performance based on academic and behavioral data. It began with an essential data cleaning phase to prepare a reliable and usable dataset. Then, an Exploratory Data Analysis (EDA) was conducted to understand variable distributions, correlations, and key characteristics.

One of the crucial steps in the data analytics process was feature engineering, where new relevant features were created to improve model accuracy and prediction quality.

Several machine learning models were tested to select the best approach. After thorough evaluation, the Random Forest model was chosen for its strong performance, ability to handle complex data, and interpretability.

A UML design was created to model the main entities and their interactions, ensuring a clear and scalable system architecture.

The developed web application offers tailored features for different users, enabling them to access their data, monitor performance, and generate personalized PDF reports.

Students:

. Access to their personal and academic data.

. View individual and class statistics.

. Personalized prediction of success or failure.

. Generate PDF reports including results ...

Teachers:

. Access overall class statistics.

. Identify struggling students and track their progress.

. Send personalized recommendations to at-risk students.

. Generate detailed PDF reports summarizing performance and personal info.

Administrators:

. Manage student, teacher, and class accounts.

. Access institution-wide indicators (enrollment, results, etc.).

. Generate institutional PDF reports summarizing key statistics.

This setup ensures effective, personalized monitoring and facilitates communication between users through alerts and recommendations.