5B 126 CHAPTER 5B ABSTRACT The prediction of running injuries is problematic. Applying machine learning techniques may be a solution. We aimed to develop a machine learning model to predict injuries in competitive runners. Twenty-three competitive runners kept a daily training log for two years. One-week (acute) and 4-week (chronic) workloads were calculated as the average training duration multiplied by the perceived exertion. The acute:chronic workload ratio (ACWR) was calculated by dividing the acute and chronic ratios. The prediction of sustaining an injury was based on the ACWR and the machine learning algorithms Bayes and Random Forest. Results show that the area under curve is low (0.43-0.60). Just as the precision of predicting an injury (0.03-0.2). Therefore, the precision of the machine learning algorithm predicting injuries must be higher to prevent running injuries actively. Keywords Human performance; predictive analysis; load; injuries; monitoring; endurance athlete
RkJQdWJsaXNoZXIy MjY0ODMw