The psychological states of athletes will influence the performance of competition. Different levels of concentration and mental representations of each sport will reflect on brainwaves. The related brainwave researches of sports´ field are applying scientific analysis to provide quantitative data, which provide researchers, coaches, and players to understand the psychological states of athletes during sports. Obviously, the performance of certain sports is related to the Alpha frequency of brainwave, the psychological states of players in different sports can be analysed with other frequencies. The proposed method in this study can be used as a prototype of high potential players¡¦ selection system in the future or can be used as adjusting the mental state for active athletes to improve training efficiency and performance. This study aims to establish classification model of sports by analysing electroencephalogram (EEG) of athletes with data mining techniques. In data collection, collects the EEG with pressure of the athletes with filling out the questionnaires in limited time. In data analysis, applies the Butterworth low-pass filter to remove noise before converts the EEG to power spectral density by Fast Fourier Transform, then extracts 24 features according to frequency bands and statistical methods as the basis of the classification model. Last but not least, constructs the classification model by machine learning algorithms such as C4.5 and CART and compares the accuracy of different classifiers. By means of proposed sports classification model coordinated with other biotechnologies such as electrocardiogram and electromyogram, which can help athletes and prospective athletes to well understand psychological conditions and choose more appropriate sports.
© Copyright 2020 Proceedings Book of The 16th ITTF Sports Science Congress Budapest (HUN), April 19th - 20th, 2019. Veröffentlicht von International Table Tennis Federation. Alle Rechte vorbehalten.