A machine learning web app built with CatBoost and Streamlit to predict the outcome of tennis matches using real-world player stats and match conditions.
👉 Try it live: https://jkwqkwz3pybyncjygrkzcq.streamlit.app/
This app predicts the winner between two professional tennis players by considering key features like:
- Player height
- Handedness (left/right)
- Current rank and ranking points
- Surface type (grass, clay, hard)
- Tournament level (e.g., Grand Slam, ATP 1000)
- Round of the match
- Best-of (3 or 5 sets)
The model is trained using historical ATP data and leverages CatBoost for accurate classification on structured features.
- CatBoost – Gradient boosting library for model training
- Pandas – For data preprocessing and feature engineering
- Streamlit – For building and deploying the interactive web app
- Python – Core programming language
- Jupyter Notebook – For exploratory data analysis and experimentation