Akademik Veri Yönetim Sistemi
Iran Journal of Computer Science, 2025 (Scopus)
This study investigates the use of machine learning techniques to analyze data from the Y-Balance test (YBT), an important tool for assessing athletes’ dynamic balance and injury risk. Precise and individualized test setups, along with strict adherence to guidelines and careful data handling, are essential for accurate measurements. YBT data were gathered from multiple sports clubs and analyzed using various machine learning models to predict outcomes and evaluate performance. The metrics such as mean squared error (MSE), root mean squared error (RMSE), and R2 were used to assess model accuracy based on individual input data. Among the models examined, gradient boosting showed the best performance with an R2 value of 0.77. These findings suggest that machine learning can effectively enhance the prediction of YBT scores, supporting the creation of tailored training programs that improve athletic performance and reduce injury risks. Specifically, XGBoost, gradient boosting, and random forest algorithms delivered the most accurate results. This methodology offers a detailed framework for experts to refine testing procedures and design personalized training regimens that boost athletes’ balance and mobility.