College students jump performance prediction based on NGA-BP neural network and the computer simulation
Dou Dan1, Feng Suqiong2
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(12C) 1117-1123
1 School of Chongqing university of posts and telecommunications, Chongqing, China
2 School of Sichuan agricultural university, Chongqing, China
In recent years, with the aggravation of the schoolwork burden, the physical quality of the student is declining. In order to encourage the college students to exercise, the State Council promulgated the “National Physical Training Standards”. In this standard, the long jump is a very important physical test project. Using the computer technology to predict the long jump performance can make the targeted training effectively for the long jump performance. In the traditional prediction methods, BP neural network is a very common method. However, in the traditional BP neural network, there exist some questions about the weight and parameter setting. In order to overcome the questions, in this paper, we propose NGA-BP neural network based on k-mean clustering. Then, we use this algorithm to predict the long jump performance for the college students with the computer simulation. The finial computer simulation shows that the method has good results.