This study aims to identify individuals who are likely to cancel in the short term with AI techniques using the dataset shared with us by the hotel.
You can examine the project in detail using the app, this app allows you to:
- Examine the aim of the project in the Description Section
- Analyze the dataset interactively with a Data Dictionary in the About Data Section
- Analyze the charts for a better understanding of data and the relationships in the Charts Section
- Examine the model performance metrics and Global Explanation of the model using SHAP in the Evaluation Section
- Make an Online and Batch prediction with Local Explanation using SHAP in the Prediction Section
streamlit_demo/
├── images/ # Images used by the app and repository
│ ├── app.png
│ ├── hotel-cancelations.png
│ └── hotel-service.png
│ ├── shap_bar_plot.png # SHAP Global Explainability Bar Plot
│ └── shap_summary_plot.png # SHAP Global Explainability Summary Plot
├── pages/ # Python files used by the app
│ ├── charts.py # Charts section
│ └── dataset.py # About Data section
│ ├── evaluation.py # Evaluation section
│ └── predict.py # Prediction section
│ ├── project.py # Description section
├── README.md # Repository description
├── explainer.pkl # SHAP Tree Explainer object
├── hotel_bookings_prepared.csv # Preprocessed dataset
├── hotel_cancellation_model.pkl # ML Model
├── requirements.txt # Project pip dependencies
├── test_features.pkl # The features used on model testing
├── welcome.py # Home Page for navigating between pages
Please feel free to contribute the ML Model or Streamlit Page.