Title: Performance prediction analysis of college aerobics course based on back propagation neural network

Authors: Jianlin Su; Hao Zheng; Yanxi Chen

Addresses: College of Physical Education, Putian University, Putian, 351100, China ' School of Physical Education and Sport Science, Fujian Normal University, Fuzhou, 35000, China ' The Department of Physical Education, Sichuan International Studies University, Chongqing, 400031, China

Abstract: To solve the problem of low accuracy of the performance prediction method of aerobics courses, the study proposes to combine the partial least squares regression (partial least squares - PLS) method with the multilayer feedforward (back propagation - BP) neural network to form a method. A new partial least squares-back propagation (PLS-BP) score prediction model, which uses the PLS method to analyse and extract the principal components, which improves the prediction accuracy of the model. It is then compared with the principal component analysis-back propagation (PCA-BP) prediction model and the partial least squares-support vector machine (PLS-SVM) prediction model. The results show that the accuracy of the PLS-BP score prediction model is 94.5%, which is better than the PCA-BP model and the PLS-SVM model. In the performance test of the constituted prediction system, the relative error value of the new score prediction system is 0.042, which has high accuracy. The experimental results show that the PLS-BP algorithm combining the PLS method and the BP neural network can improve the performance of the score prediction system, and provide a new idea for the performance improvement of the course score prediction system.

Keywords: score prediction; BP neural network; relative error value; partial least squares regression.

DOI: 10.1504/IJCISTUDIES.2023.137770

International Journal of Computational Intelligence Studies, 2023 Vol.12 No.3/4, pp.173 - 188

Received: 01 Jul 2022
Accepted: 24 Aug 2022
Published online: 05 Apr 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article