A Practical Study of Basketball Teaching Reform in Colleges and Universities Based on Big Data

• Published in: Applied mathematics and nonlinear sciences Vol. 9; no. 1
• Main Authors: Sheng, Chengjian, Lian, Chenxin, Pang, Haolin
• Format: Journal Article
• Language: English
• Published: Beirut Sciendo 01.01.2024; De Gruyter Poland
• Subjects: 97C70; Basketball; Basketball teaching reforms; Behavioral action characteristics; Big Data; Convolutional neural networks; Human multi-objective algorithms; Human posture estimation; Students
• ISSN: 2444-8656; 2444-8656
• DOI: 10.2478/amns.2023.2.01353

In this paper, the human body posture estimation algorithm is used to locate the key points of the human body in the RGB screen, and two human body multi-objective algorithms are used to predict the posture trajectory, and they can overcome the influence of the errors contained in the information recorded by the sensors to a certain extent. Secondly, the spatio-temporal graph convolutional neural network is used to identify human behavior and extract behavioral action features, and through the analysis of the action features, we understand the basketball skill level of the students and put forward the reform strategy of college basketball teaching. Sixty students from the basketball minor class at University Q’s College of Physical Education were selected as research subjects for teaching practice. The results show that the average scores of the students in spot-up shooting, half-court folding dribbling and marching one-handed over-the-shoulder shooting after the reform are higher than those before the reform by 1.80, 1.08, and 1.85, which indicates that the reform of basketball teaching based on big data can improve the students’ interest in learning and their training scores, and enhance the students’ basketball skill level.