This study investigates the effectiveness of deep learning models, including LSTM and transformer-based architectures, for predicting heart rate time series from the MIT-BIH Database. Traditional models like ARIMA and Prophet face challenges in handling noisy, sparse, and highly variable medical data, while deep learning models, particularly PatchTST, significantly outperform them across multiple metrics. The research highlights the potential of deep learning to enhance patient monitoring and cardiovascular disease (CVD) management, suggesting substantial clinical benefits. The study compares traditional models with deep learning approaches, focusing on their ability to capture complex patterns and dependencies in heart rate data. Results show that transformer-based models, such as PatchTST and iTransformer, achieve the best performance, demonstrating superior accuracy in predicting heart rate dynamics. These models effectively capture temporal dependencies and non-linear relationships, resulting in significantly lower error metrics. The study also discusses the limitations of the current research, including the small dataset and focus on heart rate alone. Future work should extend these findings to larger, more diverse datasets and real-world clinical applications to further validate and optimize model performance. The findings underscore the significant improvements that transformer-based models bring to time series forecasting of heart rate data. The study sets a benchmark for future research and practical applications in health monitoring, advocating for the adoption of advanced deep learning techniques to achieve better predictive accuracy and pave the way for more sophisticated and effective healthcare solutions.This study investigates the effectiveness of deep learning models, including LSTM and transformer-based architectures, for predicting heart rate time series from the MIT-BIH Database. Traditional models like ARIMA and Prophet face challenges in handling noisy, sparse, and highly variable medical data, while deep learning models, particularly PatchTST, significantly outperform them across multiple metrics. The research highlights the potential of deep learning to enhance patient monitoring and cardiovascular disease (CVD) management, suggesting substantial clinical benefits. The study compares traditional models with deep learning approaches, focusing on their ability to capture complex patterns and dependencies in heart rate data. Results show that transformer-based models, such as PatchTST and iTransformer, achieve the best performance, demonstrating superior accuracy in predicting heart rate dynamics. These models effectively capture temporal dependencies and non-linear relationships, resulting in significantly lower error metrics. The study also discusses the limitations of the current research, including the small dataset and focus on heart rate alone. Future work should extend these findings to larger, more diverse datasets and real-world clinical applications to further validate and optimize model performance. The findings underscore the significant improvements that transformer-based models bring to time series forecasting of heart rate data. The study sets a benchmark for future research and practical applications in health monitoring, advocating for the adoption of advanced deep learning techniques to achieve better predictive accuracy and pave the way for more sophisticated and effective healthcare solutions.