International Journal of
Information and Education Technology

Abstract

In order to improve the accuracy of sports performance prediction and solve the shortcomings of low precision and slow speed of current sports performance prediction model, this paper proposes a prediction model based on glowworm optimization neural network. Firstly, the training samples and test samples of the neural network are generated by pretreatment of the sports performance; secondly, the connection weights and thresholds of the BP neural network are determined by using the glowworm optimization algorithm, and the prediction model of the sports performance is established by learning the training samples; finally, the prediction effect is tested by specific simulation experiments. The results show that the glowworm optimized neural network improves the prediction accuracy of sports performance, and solves the limitations of other sports performance prediction models. The prediction results are more reliable, which can provide scientific decision-making basis and valuable information for sports training.