The construction of innovation system of college physical education reform based on Bp neural network

Abstract

The substance here is to workshop the current problems of physical education in higher learning institutions and construct a reform and innovation machinery of sports learning style in institutions of higher learning. This paper researches the teaching system of college sports learning style based on BP nervous enlist the services, proposes a majorization network foundation hereditary compute mode for the shortcoming that BP nervous enlist the services to be prone to land oneself in the optimal regional solution, and takes 10 influence element the results of academy sports learning style as the sample features of the data set, uses the optimized GA-BP nervous enlist the services for discipline and compares the test results with the practice price and the dispatch of other algorithms. The average error of the detection of GA-BP nervous enlist the services was 1.23, and the average error rate was 1.85%, which performed better compared with other neural networks. In terms of the poundage of the influencing factors in the export multiple, the diversity of curriculum, the advancement of teaching philosophy and the science of sports skill instruction have significant effects on the amount of sports learning style in institutions of higher learning, with the weights reaching 0.1212, 0.1387 and 0.1453, respectively. The reform and innovation of sports learning style in institutions of higher learning should develop diversified physical education courses according to students’ constitution bases and sports levels and impart scientific methods of the constitution and sports values of enjoying sports.

Keywords

• ga-bp neural network
• mean error
• samples features
• educational reform
• data set
• 97q70