Key result
The Autoregressive Model outperformed LSTM and ConvLSTM networks in predicting minute-by-minute heart rate, achieving a mean absolute error of 2.069 compared to 2.173 and 2.138 in a sample participant.
Why the study?
Heart rate time series are nonlinear and nonstationary, yet accurate modeling and reliable prediction are important to identify underlying cardiovascular diseases and prevent conditions.
Does an Autoregressive Model improve heart rate prediction compared to deep learning models in heterogeneous participants?
Comparison
Autoregressive Model vs Long Short-Term Memory Network vs Convolutional Long Short-Term Memory Network
Design
Comparative modeling and forecasting study
Follow-up
10 days
Authors
Loading...
Should not change clinical heart rate monitoring; leaves open validation across heterogeneous participants.
Does an Autoregressive Model improve heart rate prediction compared to deep learning models in heterogeneous participants?
Absolute Event Rate: 2.069% vs 2.173%
Minute-by-minute heart rate prediction can be accurately performed using a linear Autoregressive Model, which outperformed more complex deep learning architectures.
Staffini et al. (2021) studied this question. Autoregressive Model vs. Long Short-Term Memory Network and Convolutional Long Short-Term Memory Network was evaluated on Mean absolute error and root mean square error for heart rate prediction. The Autoregressive Model outperformed LSTM and ConvLSTM networks in predicting minute-by-minute heart rate, achieving a mean absolute error of 2.069 compared to 2.173 and 2.138 in a sample participant.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: