🎓 Student Performance Prediction System

A machine learning web application that predicts student exam scores based on various academic and personal factors. Built with Streamlit and Random Forest Regression.

Developed by: Ahmad Omar

📊 Dataset

This project uses a merged student performance dataset from Zenodo.

Dataset Information:

• Total Records: 14,003 students
• Total Features: 14 features
• Target Variable: Exam Score (0-100 scale)
• File Format: CSV (merged_dataset.csv)

🎯 What is Exam Score?

The model predicts a student's Exam Score on a scale of 0-100, representing their expected performance on an exam.

Exam Score Scale Interpretation:

• 90-100 = Excellent (نایاب)
• 80-89 = Very Good (زۆر باش)
• 70-79 = Good (باش)
• 60-69 = Average (مامناوەند)
• Below 60 = Needs Improvement (پێویستی بە باشترکردن هەیە)

Pass/Fail Criteria:

• 60 or above = Pass (تێپەڕ)
• Below 60 = Fail (شکست)

🔍 Features (14 Factors)

The model uses 14 factors to predict student performance, organized into the following categories:

👤 Demographic Information (4 factors)

1. Gender - Student's gender (Male/Female)
2. Age - Student's age
3. Learning Style - Preferred learning method (Visual, Auditory, Kinesthetic, Reading/Writing)
4. Motivation - Motivation level (Low, Medium, High)

📖 Study Behaviors & Engagement (6 factors)

5. Study Hours - Hours studied per week
6. Attendance - Attendance rate (%)
7. Assignment Completion - Assignment completion rate (%)
8. Online Courses - Number of online courses taken
9. Discussions - Participation in discussions (Yes/No)
10. Extracurricular - Engagement in extracurricular activities (Yes/No)

💻 Resources & Technology (4 factors)

11. Resources - Resource access level (Low, Medium, High)
12. Internet - Internet access availability (Yes/No)
13. EduTech - Use of educational technology (Yes/No)
14. Stress Level - Stress level (Low, Medium, High)

🌍 Language Support

The application supports two languages:

• 🇬🇧 English - Full English interface
• 🇹🇯 کوردی (Kurdish) - Complete Kurdish translation with RTL support

Users can switch between languages using the navigation buttons at the top of the page.

🚀 Installation

1. Clone the repository

   git clone https://github.com/a7x3a/StudentPerformanceApp.git
   cd StudentPerformanceApp

2. Create a virtual environment

   python -m venv venv

3. Activate virtual environment

   • Windows:

     venv\Scripts\activate

   • Linux/Mac:

     source venv/bin/activate

4. Install dependencies

   pip install -r requirements.txt

5. Download the dataset

   • Ensure merged_dataset.csv is in the project root directory
   • The dataset should contain 14 feature columns plus ExamScore and FinalGrade

📝 Usage

1. Train the Model

python train.py

This will:

• Load the dataset from merged_dataset.csv
• Train a Random Forest Regressor
• Save the model as model.pkl
• Display training and testing R² scores

Note: All features in the dataset are already numeric, so no encoding is required.

2. Run the Web Application

streamlit run app.py

The application will open in your default web browser at http://localhost:8501

3. Make Predictions

1. Select your preferred language (English or Kurdish)
2. Fill in all the required fields in the form:
   • Demographic Information (Gender, Age, Learning Style, Motivation)
   • Study Behaviors & Engagement (Study Hours, Attendance, Assignments, etc.)
   • Resources & Technology (Resources, Internet, EduTech, Stress Level)
3. Click "🔮 Predict Exam Score" (or "🔮 پێشبینی نمرەی تاقیکردنەوە" in Kurdish)
4. View the predicted exam score (0-100) along with performance level and pass/fail status

📁 Project Structure

StudentPerformanceApp/
│
├── app.py                          # Streamlit web application
├── train.py                        # Model training script
├── model.pkl                       # Trained Random Forest model
├── merged_dataset.csv              # Dataset file (CSV format)
├── requirements.txt                # Python dependencies
├── README.md                       # This file
└── venv/                           # Virtual environment (not in git)

🛠️ Technologies Used

• Python 3.x
• Streamlit - Web application framework
• Pandas - Data manipulation
• Scikit-learn - Machine learning (Random Forest Regressor)
• Joblib - Model serialization

🙏 Acknowledgments

• Developer: Ahmad Omar (a7x3a.dev)

---

Note: Make sure the merged_dataset.csv file is in the project directory and the model has been trained (model.pkl exists) before running the application.