Student Writing Score Predictor

This is a Streamlit web application that predicts a student's writing score based on their math and reading scores using a Linear Regression model trained on the Students Performance Dataset.

Assumption: If a student performs well in math and reading, they will likely perform similarly in writing.

Live Demo

🔗 Launch App on Streamlit https://studentscorepredictor-83dblgddhbjzylyez8sgoe.streamlit.app/

Features

• Predict Writing Score from Math and Reading Scores

• Shows Model Accuracy (R² Score) after prediction

• Clean UI with interactive sliders

• Uses Linear Regression for continuous prediction

• Model is saved using pickle and re-used in production

• Built with Streamlit, scikit-learn, pandas, NumPy, Matplotlib

Project Structure

student-score-app/
├── app.py                  ← Streamlit app
├── train_model.py          ← One-time training script
├── StudentsPerformance.csv ← Dataset
├── model.pkl               ← Pickled trained model
├── X_test.pkl              ← Test data for accuracy display
├── y_test.pkl
├── requirements.txt        ← Python dependencies
└── README.md               ← You are here

How It Works

1. Dataset

Based on the StudentsPerformance.csv dataset containing:

• Categorical: gender, race/ethnicity, parental education, lunch, test preparation

• Numerical: math, reading, writing scores

2. Model Assumption

We assume writing score depends mostly on math and reading scores. So:

• Features: math_score, reading_score

• Target: writing_score

3. Model Training (train_model.py)

• Encodes categorical variables

• Normalizes numerical features

• Trains a LinearRegression model

• Saves:

  • model.pkl

  • X_test.pkl, y_test.pkl (for calculating R² in production)

4. Streamlit App (app.py)

• Loads model and test data

• Takes input using sliders

• Predicts writing score

• Displays model performance (R² Score)

Example

_ _

Installation & Run Locally

git clone  https://github.com/MansoobeZahra/Student_score_Predictor.git
cd student-score-app
pip install -r requirements.txt

# First: run training script
python train_model.py

# Then: run Streamlit app
streamlit run app.py

Requirements

pandas
numpy
scikit-learn
streamlit

Author

Mansoob E Zehra | Pakistan | Computer Science Student| GitHub

License

This project is open-source and available under the MIT License.