Grey BP neural network-based women heptathlon performance prediction application

摘要

The paper makes statistics of world women heptathlon annual best performances during 2005????2013, takes prediction on heptathlon best performance in 2013 as research objects. According to statistical world women heptathlon annual best performance during 2005????2012, the paper establishes sports performance prediction model that is series combined by GM(1, 1) grey prediction model and BP neural network prediction model, and applies established model to predict world women heptathlon annual best performances in 2013. By comparing world women heptathlon annual best performance actual value with model predicted value in 2013, study and analyze grey BP neural network prediction model application in sports performance prediction problem. Research results show that for sports performance prediction problem, grey BP neural network prediction model’s prediction accuracy is high, application is simple, and generalizability is strong, it has more advantages over single GM(1, 1) grey prediction model and BP neural network model by comparing.