Application of optimization of teacher teaching paths in physical education based on GCN

With the advancement of global education reform, the role of physical education in cultivating students' physical and mental health and improving their comprehensive quality has become increasingly prominent. However, the traditional physical education model faces many challenges, such as single teaching methods, lagging information processing, and it is difficult for conventional methods to fully considering the interactive relationship between students. It lacks dynamic adjustment capabilities and is difficult to adapt to students' personalized needs and changes. Therefore, this paper introduces a deep learning method based on graph convolutional networks (GCN) to model the interactive relationship between students. Secondly, the recurrent neural network (RNN) model is used to further analyze and model the students' physical learning situation, and the students' future performance is predicted based on the time series. Finally, the Q-learning model is used to optimize teaching decisions through students' feedback and dynamically adjust the teaching path. Experiments show that the teaching path provided by the research method in this paper has a positive effect on students' athletic ability. The average scores of students across six test items have shown improvement. Meanwhile, the accuracy of teaching path optimization in the experiment exceeds 0.8, with precision surpassing 0.85. Additionally, both recall and F1 score are exceptional. The matching analysis of students' learning progress and learning goals showed that the learning progress of most students was within expectations. The personalized path studied in this paper can make good use of most students and conduct differential analysis on students of different ability levels. Compared with the benchmark method and traditional method, the research method in this paper shows obvious advantages, further confirming the positive impact of the personalized teaching path on students' learning outcomes, offering strong theoretical support and a practical foundation for applying intelligent teaching in future physical education.