Student Performance Prediction

Regression of an exam score (0-100) from study hours, previous score, sleep hours, practice papers, attendance and extracurricular activity.

Features

• Deterministic synthetic dataset, no downloads
• Data-leak-safe workflow: the train/test split happens BEFORE any scaling or encoding, and all transforms live inside a Pipeline fit on the training split only
• Random Forest regressor
• Extracurricular activities accepted as 1/0/yes/no in the CLI and API
• CLI and Flask web app (form + JSON API)

CLI input order

hours_studied, previous_score, sleep_hours, sample_papers_practiced, attendance_percent, extracurricular_activities

Quickstart

pip install -r requirements.txt

python -m studentperformance train
python -m studentperformance predict 7 85 7 3 90 yes

Output:

score=93.0

Run the web app:

python -m studentperformance.app

Open http://127.0.0.1:5000 or POST JSON to /api/predict.

Tests

python -m pytest -q

Project Structure

student-performance-prediction/
  studentperformance/
    data.py      # synthetic dataset generation
    model.py     # split-then-transform pipeline, training, prediction
    cli.py       # train / predict commands
    app.py       # Flask app
  tests/test_smoke.py
  requirements.txt

License

MIT