Applications of continual construction of sports training system based on data mining technology

Abstract

Objective: In competitive sports, one of the most effective ways to improve athletes' competitive level is intensive training. The current training pro-cess is mainly based on the coach’s existing experience and the athlete’s personal physical condition to formulate a training plan, but the formulation of this training plan obviously lacks scientific basis. This paper combines data mining technology and image recognition tech-nology to construct a sports training system. In order to better deal with the over-dispersed and multi-peak count data, this paper proposes a mixed integer generalized autoregressive conditional heteroscedasticity model. The model uses virtual reality technology to construct train-ing scenes to improve athletes' training immersion, and can use digital technology to identify sports training processes, correct actions, and improve sports training effects. Finally, this article combines with experimental research to verify the training effect of the system con-structed in this article.