Abstrakt

Aerobics performance evaluation optimization model research based on bp neural network

Dan Li, Fenglong Sun

In recent years, aerobics have been rapidly developed in China; these achievements are also obvious to all. To further improve Chinese aerobics level, it is indispensible to make analysis of its result. Each itemevaluation indicators which require being analyzed and defined exactly include many kinds of indicators; it shows each indicator performance and overall performance relations by establishing three layersÂ’BP neural networkwhich contains input and output layer as well as the hidden layer. The paper applies mass and athletesÂ’ two kind of people performances to make comprehensive evaluation for aerobics performance prediction, the result shows each sample valueÂ’s actual value and predicted value are very approximate, the former error is less than 5%, and the later error is less than 0.1, which indicates BP neural network has very high and widely fitness to aerobics comprehensive evaluation after training