Feature extraction and personalized scheme generation of college students’ physical exercise behavior driven by deep learning
Year: 2026
Type: article
Abstract: Under the background of the continuous progress of digital and intelligent physical education teaching in colleges and universities, it has become a widely concerned research topic in the field of sports research to accurately identify the behavior characteristics of college students’ physical exercise with relevant data and carry out personalized intervention. This study takes college students’ physical exercise behavior as the core concern, and tries to construct a method to extract behavior characteristics and generate personalized exercise programs by combining deep learning technology. With the help of the collection of multidimensional sports behavior data, the characteristics of college students in sports frequency, intensity and persistence are modeled and analyzed, and a behavior recognition model combining convolutional neural network with bidirectional long-term and short-term memory network is established to describe the dynamic change law of individual sports behavior. By means of attention mechanism, the key behavioral characteristics are expressed in a weighted way, so that a personalized exercise program with strong adaptability can be further generated. The final result of the experiment shows that this method has a good performance in the accuracy and stability of behavior recognition and the actual effect of personalized intervention, which can effectively improve the participation and sustainability of college students’ sports. The results of this study suggest that the proposed deep learning-based approach shows promising performance in behavior recognition and personalized intervention within the current sample, providing a feasible reference for further validation and application in college physical education settings.
Source: Discover Artificial Intelligence
Author: Jisheng Wang
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: gold
Estimated APC paid: $1,590
OpenAlex ID: W7204237963