Construction of a neural network model for performance prediction in shot put athletes

This paper is to investigate the role of artificial neural network models in the prediction of sports performance and to establish artificial neural network models to evaluate the correlation between athletes’ special performance and special physical quality. The results show that the artificial neural network model overcomes the shortcomings of multiple regression models and gray models that require a pre-determined mathematical model, and more accurately reflects the functional relationship between the training indexes of the special quality of shot put athletes and the special athletic performance, and more accurately predicts the special performance of shot put athletes. It further showed that the special physical quality of shot put athletes is an important basis for their special athletic ability. With the physical quality index as the independent variable and the special performance as the dependent variable, the prediction model of the athletes’ special performance can be established, which can accurately diagnose and evaluate the development level of the athletes’ physical quality, and clarify the key points and objectives of the training content, so as to improve the scientific level of the training of shot put.