Investigation of a Sports Teaching Performance Evaluation Model Based on Decision Tree Algorithm

Different factors have great influence on students' physical education scores, and the existing evaluation methods have some problems such as confusion and inaccurate prediction. To address these challenges, this paper proposes a physical education teaching achievement evaluation model based on the decision tree algorithm, encompassing four key steps: data acquisition, feature selection, decision tree construction, and model evaluation. The C4.5 algorithm is employed to build and optimize the decision tree model. In the final evaluation stage, cross-validation and confusion matrix methods are utilized to assess the model's validity and reliability. Compared to traditional manual data processing methods, the use of the decision tree algorithm significantly enhances prediction accuracy. This evaluation model provides a valuable reference for assessing physical education teaching outcomes and guiding teaching improvements. Additionally, the model's decision rules facilitate the design and development of a curriculum grade prediction system, enabling teachers to monitor students' learning progress and predict course grades in real-time.