We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
A framework for applying the Logistic Regression model to obtain predictive analytics for tennis matches.
- Authors
Friligkos, Georgios; Papaioannou, Evi; Kaklamanis, Christos
- Abstract
In this work, we apply the Logistic Regression (LR) model for predicting the outcome of tennis matches and providing win probability for participating/competing players. Prediction models are classified as machine learning methods, which generate predictions for future scenaria exploiting existing data collections. With the objective to maximize the accuracy of our predictions, we apply the Logistic Regression model under various parameter configurations seeking for an optimal combination of independent variables (features). We present and discuss promising results obtained via the holdout-validation method and the cross-validation method. Furthermore, as a proof of concept of our approach, we designed and implemented a web platform, where users can obtain real-time predictions for tennis matches. Our LR-based framework combined with Artificial Intelligence (AI) can serve as a paradigm in the field of sports analysis for predicting outcomes and evaluating athletic performance.
- Subjects
ATHLETIC ability; LOGISTIC regression analysis; ARTIFICIAL intelligence; PYTHON programming language; TENNIS tournaments; CONFIGURATION management
- Publication
Technium, 2023, Vol 15, p60
- ISSN
2668-778X
- Publication type
Article
- DOI
10.47577/technium.v15i.9616