Athletes' performance and injury management in sports training using association rules and data mining techniques

Optimizing athletic performance while minimizing injury risk remains a central challenge in sports science. Although large volumes of athlete monitoring data are now available, existing studies primarily rely on modeling or isolated statistical analyses, with limited exploration of interpretable pattern discovery for understanding multi-factor relationships. This study addresses this gap by applying association rule mining techniques, including Apriori, FP-Growth, and Eclat, to identify conditional relationships within a comprehensive sports training dataset comprising demographic, physiological, psychological, and training-related variables. Unlike prior work focused mainly on classification accuracy, this approach emphasizes interpretable rule extraction to identify key combinations of training intensity, recovery status, sleep, and prior injury history associated with performance and injury outcomes. Rule quality is evaluated using standard association measures, while validation is conducted through classification-based metrics. The results show that association rule mining can reveal stable and high-impact patterns that may inform monitoring interpretation, support hypothesis generation, and help prioritize factors for future prospective validation. This study highlights the value of interpretable data mining frameworks for exploring association structures in sports training data within a non-causal analytical setting.