Youth Sports Special Skills’ Training and Evaluation System Based on Machine Learning

Nowadays, most sports training adopts such a mode, that is, the coach guides and observes the movements of the trainer on the side, or uses the camera to collect the data information in the training video. After the training, the trainer needs to experience and comprehend it according to the coach’s guidance and suggestions, which leads to a poor training effect to a certain extent, and the training movements are not so standardized. The most important thing is that the physical condition of the trainer will change at any time during the exercise. However, the traditional training mode cannot perceive and predict it at all, and it may even lead to the training volume being too large for the athletes to bear, which also causes irreversible effects on the body. Based on this, this paper aims to use the relevant theories and technologies of machine learning to build a system for the training and evaluation of sports-specific skills. This paper automatically determines and guides the movements of the trainer, which greatly improves the training efficiency. The movements are also more standardized, and the workload of the coaches is also reduced. In addition, the system includes explained functions. It uses holographic projection to play various videos of training and guidance, which brings convenience to trainers to observe and understand specific sports behaviors. Finally, the data results of the experimental subjects were analyzed to verify the effectiveness and feasibility of this training evaluation system, in which their special scores increased by 1.51% compared with the previous ones.