& 151 SUMMARY objective was to test whether traditional statistical analysis or machine learning applied to the inherently incomplete dataset provides better insights and predictions. We found a statistical relationship between workload and injury risk. However, both statistics and machine learning are limited in delivering actionable predictions to prevent injuries. Conclusion This thesis showed that personalised data, machine learning, sensitive performance indicators, advanced statistics, and a causal roadmap in combination with a causal model could help to reduce the data analytics gap. As a result, this creates the possibility of extracting meaningful insights and predictions from physical activity and physical performance data. While reducing the data analytics gap, we showed the potential of data analytics to gain meaningful insights and predictions on physical activity, physical performance, or injuries, enabling more informed interventions in physical activity. These results provide a foundation for future research to reduce the data analytics gap even more and potentially the realisation of automated monitoring, prediction, and coaching systems.
RkJQdWJsaXNoZXIy MjY0ODMw