One-sentence statement: Wearable computing systems and data mining concepts have the potential to support biomechanical research outside the lab and to help adapt sports research results to the individual. Wearable computing systems (Fig. 1, blue arrows) play an increasingly important role in recreational and elite sports. They comprise of two parts. First, sensors for physiological (ECG, EMG, …) and biomechanical (accelerometer, gyroscope, …) data recording are embedded into clothes and equipment. Second, embedded microprocessors (e.g. in smartphones) are used for monitoring and analysis of the recorded data. Together, these wearable computing systems can provide real-time information and feedback for scientific studies in the real sports situation (i.e., outside the lab). Data mining concepts (Fig. 2, red arrows) provide a toolbox for the analysis of the considerable amount of physiological and biomechanical data that can be generated in sports science studies. Especially in studies using wearable computing systems, the number of participants and variety of measured data is unlimited in general. Traditional statistical analysis methods commonly cannot handle this amount of data easily. Thus, the analysis is often restricted to individual variables rather than multidimensional dependencies and a considerable amount of information is neglected. Moreover, the results are frequently biased by the expectation of the researcher, and are furthermore group based rather than giving information about the individual. Here, the objective, data-driven methods from data mining can contribute by offering useful tools for the analysis tasks. These tools have the ability to deal with large data sets, to analyze multiple dimensions simultaneously, to work data-driven rather than hypothesis-driven, and to provide Tiresults that are tailored to the individual.
© Copyright 2014 International Calgary Running Symposium, August 14-17, 2014. Alle Rechte vorbehalten.