Enhancing Physical Education: A Dynamic Fuzzy Neural Network-Based Information Processing System Design

With the advent of digital intelligent education, educational resources are expanding. In the realm of physical education, data encompassing students’ physical fitness test results, sports performance, and health status stand as pivotal pillars for the assessment of teaching. Consequently, this manuscript conceives a dynamic fuzzy neural network-based system for processing physical education information. Primarily, this work introduces compensatory fuzzy neurons into the Dynamic Fuzzy Neural Network (DFNN) and advances the creation of a Generalized Dynamic Fuzzy Neural Network (GDFNN). Furthermore, the GDFNN is seamlessly integrated with the Reinforcement Learning (RL) method to devise a neural network inverse controller equipped with online learning capability, employing the temporal difference learning method within RL. The empirical findings demonstrate that the enhanced generalized dynamic fuzzy neural network attains impressive results, yielding 0.0025, 0.2668, and 0.9356 on the root mean square error, standard root mean square error, and homogeneity coefficient, respectively. These outcomes are juxtaposed with those of eight contemporary algorithms, including the RBF method, BP algorithm, and other prevalent feedforward neural network algorithms. The root mean square error (RMSE), standard Root Mean Square Error (RRMSE), and Equality Coefficient (EQU) performance metrics register an augmentation of 9%, 11.1%, and 6.7%, respectively. This enhancement signifies a substantial boost in predictive efficiency, thereby effectively advancing the intelligent development of the information processing system for physical education.