Analyzing Golf Performance Metrics Using TrackMan Data

This capstone project used TrackMan radar data from an amateur golfer to develop a personalized, data-driven approach to improving swing performance. The objective was to predict carry distance using key swing metrics such as club speed, ball speed, launch angle, spin rate, and club type. Three machine learning models, Linear Regression, Random Forest, and XGBoost, were built and evaluated, with XGBoost achieving the highest accuracy. The study also applied global optimization techniques like simulated annealing, dual annealing, and Bayesian optimization to enhance feature selection, model tuning, and performance. SHAP analysis and feature importance scores revealed that ball speed was the most critical predictor of carry distance. The final model delivered actionable recommendations for improving swing consistency, optimizing training, and selecting the appropriate clubs. This case study demonstrates the power of data science in sports analytics and highlights how machine learning can drive performance improvements even at the recreational level.