This master's thesis focuses on the automation of training unit evaluations for professional athletes, particularly middle and long-distance runners. The main problem is the lack of automated training unit evaluations, which leads to excessive manual work and less efficient training diary management. The aim of the thesis is to propose a method for evaluating training units from data generated by Garmin sports watches, which takes into account key general training indicators (OTU) and speci... zobrazit celý abstraktThis master's thesis focuses on the automation of training unit evaluations for professional athletes, particularly middle and long-distance runners. The main problem is the lack of automated training unit evaluations, which leads to excessive manual work and less efficient training diary management. The aim of the thesis is to propose a method for evaluating training units from data generated by Garmin sports watches, which takes into account key general training indicators (OTU) and special training indicators (STU). The first five steps of the CRISP-DM methodology were used in this work: understanding the context, understanding the data, data preparation, modeling and evaluation. To achieve the goal, qualitative research was conducted using interviews with professional runners, analysis of available software tools and processing data from Garmin watches. Subsequently, predictive models were created for predicting the type and difficulty of training. The results of the work show that training unit evaluations can be automated using hard metrics of OTU and STU, simplifying the process of evaluating training diaries for athletes and coaches. However, there was not enough data available for evaluation using RPE (Rating of Perceived Exertion) for comparison, as Czech elite athletes do not use this metric. This thesis thus contributes to better evaluation of training cycles, which can have a positive impact on athletes' training processes and goal achievement. |