A Multi-level Fusion System for Intelligent Capture and Assessment of Student Activity in Physical Training based on Machine Learning

 

 Mustafa Altaee1, A. Jawad 2, Mohammed Abdul Jalil3,4, Sanaa Al-Kikani5, Ahmed Oleiwi6, Hatıra Günerhan7

 

1Department of medical instruments engineering techniques, Alfarahidi University, Baghdad, Iraq

2Computer Communications Engineering Department, National University of science and technology , Thi Qar, Iraq

3 Department of Computer Engineering techniques, Alturath University college, Baghdad, Iraq

 4MEU Research Unit, Middle East University, Amman 11831, Jordan.

5 Department of  Physical Education and Sport Science, Al Mustaqbal University College, 51001 Hilla, Babylon, Iraq

6Biomedical Engineering, College of Engineering, University of Warith Al-Anbiyaa , Karbala, Iraq 

7 Department of Mathematics, Faculty of Education, Kafkas University, Kars, Turkey;

 

Emails: m.altaee@alfarahidiuc.edu.iq; a.jawad@nust.edu.iq; mohammedjalil24@gmail.com; Dr.Sanaa@uomus.edu.iq; A.oleiwi@uowa.edu.iq;

hatira.gunerhan@kafkas.edu.tr

 

 

Abstract

 

To record and evaluate students' physical education class participation, this study proposes using a Machine Learning aided Physical Training Framework (ML-PTF). Improve student achievement in physical education with the help of the Multi-level Fusion System that employs machine learning strategies. The system integrates sensor data, video data, and contextual data to deliver a holistic and precise evaluation of student engagement. This study's simulation analysis shows that the ML-PTF improves the reliability of evaluating universities' physical education programs. A important reference path and paradigm for advancing tertiary-level physical education for graduates, the multi-level fusion system also provides an investigation of information technology and language education integration. The experimental findings demonstrate that the ML-PTF is superior to other approaches in terms of learning rate, f1-score, precision, and probability, as well as student engagement, involvement, and recognition accuracy.

 

Keywords: Physical Education;  Machine learning; A Multi-level Fusion System; Assessment model; Student Activity Prediction.