How muhammadmoaz/st1capstoneproj works
This is a country-continent prediction project: a training script explores country GDP and population data, trains and evaluates classifiers, and saves an XGBoost model. Users can enter those features in either a Flask web form or a Tkinter desktop tool and receive a predicted continent. The source excerpts show these paths and the model artifact; the template contents and CSV schema were not sampled.
Main entry points: WebTool.py Web predictor · best_model.pkl Trained model · main.py Data exploration · index.html Prediction form
Components 9
Model training
- Data exploration
main.py - Feature preparation
main.py - Model evaluation
main.py - Trained model
best_model.pkl
Prediction interfaces
- Web predictor
WebTool.py - Prediction form
templates/index.html - Desktop predictor
DesktopTool.py
Other components
- User
- Country dataset
countries.csv
Connections 10
- Country dataset → Data exploration: read by
main.py - Data exploration → Feature preparation: feeds
main.py - Feature preparation → Model evaluation: trains and tests
main.py - Model evaluation → Trained model: saves
main.py - User → Web predictor: submits features
WebTool.py - Web predictor → Prediction form: renders
WebTool.py - Prediction form → Web predictor: posts features
WebTool.py - Web predictor → Trained model: loads and predicts
WebTool.py - User → Desktop predictor: enters features
DesktopTool.py - Desktop predictor → Trained model: loads and predicts
DesktopTool.py