The Prediction of Heart Rate during Running Using Bayesian Combined Predictor

Heart rate is an important physiological sign that can reflect the human body burden and exercise intensity. Due to the sensor faults, energy saving requirement and other factors, heart rate prediction become very important for the monitoring of human body during exercise. Most of the available researches using single model for prediction of heart rate have a bad performance with a greater prediction error. In light of this, we formalize a multi-step prediction scheme for heart rate during running using the Bayesian combined predictor. We first construct a Neural network predictor and a Linear regression predictor as the basic prediction models by parameter learning with training data, and then construct a Bayesian combined predictor by training the weights of the basic predictors to give the multi-step prediction process of heart rate during running. Finally, we evaluate the performance of the proposed scheme using actual measurement data of wearable devices of several runners, and the simulation results tell us that the proposed scheme can achieve better performance of prediction error compared to the available schemes.