Tennis AI to predict the winner in ATP matches with 3 different models and a web UI
Overview
Tennis AI is an innovative machine learning-driven system designed to predict tennis match outcomes with impressive accuracy. By leveraging extensive player statistics, historical performances, and various match conditions, it offers users a reliable tool to gain insights into potential match results. Whether you’re an avid tennis fan or a bettor looking for an edge, Tennis AI presents compelling analytics to enhance your experience and understanding of the game.
What sets Tennis AI apart is its robust analytical capabilities. Employing advanced machine learning models, it not only forecasts match outcomes but also provides confidence scores, helping users gauge the reliability of each prediction. The modern web interface is intuitive and responsive, making it easy to access valuable data and predictions right at your fingertips.
Features
- Advanced ML Models: Three powerful models (CatBoost, XGBoost, Random Forest) ensure accurate match predictions, with CatBoost being the default choice for reliability.
- Player Database: Access a comprehensive collection of player statistics and historical data to inform your predictions.
- Smart Search: Effortlessly find players by name or country code, making it easy to locate specific information.
- Prediction Interface: Get match predictions along with confidence scores and winning odds, giving you insights into likely outcomes.
- Modern Web Interface: A clean and responsive design built with SvelteKit that enhances the user experience across all devices.
- REST API: FastAPI backend with multiple endpoints allows for quick access to predictions and player data.
- Feature Engineering: The models utilize 137 carefully engineered features, including player characteristics and match context, for in-depth analysis.
- Dataset Coverage: Features data from tennis matches up to and including the entire 2024 season, ensuring the predictions are based on the latest statistics.