A Study of Influencing Factors and Decision Making for Race Momentum Based on Transformer and Decision Tree Modeling

In this paper, we explore the key factors of athlete performance analysis and decision support using deep learning Transformer model and decision tree model. By comprehensively evaluating momentum fluctuations and investigating their correlation with score dynamics, we reveal the impact of skill level, fatigue index, and psychological factors on match outcomes. Combined with logistic regression modeling, we successfully predicted fluctuations during the match and found that the fatigue index may have a negative impact on the athlete's swing. Further decision tree analysis showed that psychological factors and skill level had a significant role in influencing game performance. Finally, key recommendations for athletes are presented, aiming to help them achieve better performance in future competitions. This study provides an important reference value for a deeper understanding of momentum and decision making in sports, thus helping players and management teams to make more informed decisions.