Research on Heartbeat Temporal Signal Recognition Based on Machine Learning and Deep Learning

The rapid development of AI technologies, such as machine learning and deep learning, has provided new directions for medical research. With extensive data mining and learning, AI has demonstrated strong potential in tasks like heartbeat signal recognition and image diagnosis. This study utilizes real heartbeat signal samples from hundreds of thousands of instances to research heartbeat signal prediction using various machine learning and deep learning algorithms, including LightGBM, XGBoost, and CNNs. The results indicate that two CNN models, CNN_1 and CNN_2, achieved accuracies of 99.10% and 99.21% on the test set, significantly outperforming XGBoost's 98.79% and LightGBM's 98.97%. Through further data optimization and model fusion, under a weighted average fusion strategy, the evaluation metric for the CNN models was improved to a maximum of 99.53%. The experimental results of this paper provide data support and a feasible solution path for the auxiliary diagnosis of medical big data.