Abstract

Predict the neural network mathematical model of basketball team scores based on improved BP algorithm

Author(s): Chengliang Li

Sports performance has increasingly become the measure and the symbol of a countryÂ’s economic strength, educational level, and comprehensive national strength. As one of the three big ball games, basketball also occupies a pivotal position in the national sport. In order to achieve better play results and provide a good guide for peacetime training and policy development, the study applies the improved BP neural network to establish the mathematical prediction model of the menÂ’s basketball performance. Based on MATLAB mathematical software, according to the actual data as shooting rate, three-point shooting, assists statistics, rebounding, obtained from the 2004 Olympic Games, the 14th World Basketball Championships, the 2006 Intercontinental Cup basketball game and the 2004Athens Olympic Games, the 2012 London Olympic Games sports scores were predicted. The analysis of the predicted results and the actual results showed that the error is small; there is a theoretical feasibility of the algorithmin practical problems.After constantly improving and updating the internal data of the model, the model can provide better service for basketball development and can be applied to other areas.