About Our Project

Predicting Student Dropout Risks Using Educational Data

Overview

In the ever-evolving educational landscape, student retention is a critical focus for educators and administrators. Our innovative solution addresses this challenge by employing artificial intelligence to predict the likelihood of students leaving school. This proactive strategy enables educators to provide timely interventions, supporting at-risk students effectively.

The Challenge

With a wealth of data available, this application leverages various datasets to accurately forecast dropout risks. It analyzes:

• Class Attendance: Records of student presence and absences.
• Academic Performance: Exam scores, assignment grades, and overall performance metrics.
• Disciplinary Records: Logs of any disciplinary actions or incidents.
• Extracurricular Activities: Details on participation in clubs, sports, and other activities.
• Personal Information: Insights into age, gender, socioeconomic status, and more.

Why It Matters

• Early Intervention: Early detection of at-risk students allows schools to offer supportive mechanisms, enhancing student outcomes.
• Data-Driven Decisions: A data-centric approach optimizes decision-making and resource allocation.
• Holistic Support: Considering a broad range of factors, the model addresses individual student needs comprehensively.

Join Us in Advancing Education

By predicting dropout risks, our model helps create environments where every student has the opportunity to succeed. Discover the difference our model can make in your institution today.

Application Overview

This Flask application predicts the likelihood of a student dropping out using several input factors. It relies on a machine learning model (Logistic Regression) to predict dropout probabilities based on features extracted from the data.

Key Features

Data Preparation:

• Loads student data from a CSV file.
• Calculates linear scores for dropout probability using a weighted sum of features.
• Normalizes dropout probability and applies thresholds to categorize students.

Features and Weights:

• Utilizes features like average grade, age, gender, economic status, attendance, discipline incidents, and extracurricular activities.
• Defined weights reflect the impact of each feature, indicating positive or negative influence on dropout risk.

Machine Learning Model:

• Logistic Regression: Predicts dropout likelihood using labeled data.
• Data Scaling: Applies StandardScaler for normalizing data inputs.

Flask Application:

• Routes:
  • /: Displays an index page for user input.
  • /predict: Processes form submissions to predict dropout probability and renders the result.
• CORS Support: Enables cross-origin requests compatibility.

Prediction Logic:

• A prediction function scales new inputs and estimates dropout probability using the logistic regression model, converting results to a percentage for ease of understanding.

User Interface:

• Index Page: User-friendly form for data input.
• Result Page: Displays dropout probability and risk level guidance.