Understanding the Effects on Balance for Elite Platform Divers Using Machine Learning

Machine learning plays a crucial role in our society’s efforts to combat injury to athletes. The purpose of this paper is to analyze and examine the studies conducted and to predict the condition/trial type (eyes open, closed, etc.) based on anterior-posterior sway (AP sway) and medial-lateral sway (ML sway) from the dataset provided by Rachel McCormick on the University of North Carolina Wilmington’s diving team. In McCormick’s experiment, a force plate is used to measure a series of trials conducted. It is surmised that the more dives performed, the worse the individual’s balance became over time. This work uses regression, deep learning, and classification to predict AP sway, ML sway, trial type, and subject number. We were able to accurately predict all features with over 90% accuracy. This means that not only are we able to accurately predict the amount a participant would sway in the anterior-posterior and medial-lateral direction, but also predict who the participant was, and the manner in which they were standing on the force plate.