Deep Learning-Based Prediction Model for Exercise Heart Rate Variability

The purpose of this study is to develop a deep learning-based sports heart rate variability prediction model to improve the accuracy and real-time prediction of heart rate variability. In this study, a multi-layer neural network structure was used, combined with a convolutional neural network and a long short-term memory network, to process and analyze the collected heart rate data during exercise. Through comparative experiments with traditional prediction methods, the prediction accuracy of the model is improved by $15 \%$, the real-time response speed is increased by $30 \%$, and the parameters can be adjusted adaptively according to individual differences. The results show that the proposed deep learning model can effectively capture the dynamic change characteristics of heart rate variability during exercise, and provide reliable technical support for athlete training monitoring and personalized training program formulation.