Glucose Level Prediction Based on Data Driven Method

Diabetes is a chronic disease affecting a large number of human population worldwide. Accurate prediction of blood glucose plays an important role for diabetic patients to control the blood glucose in the normal range. In this paper, we use four popular data driven prediction methods for multi-steps ahead prediction with only the historical glucose values as input. Moreover, experiments are carried out to gain insight of the forecast delay phenomenon in the prediction. The reasons leading to the prediction delay are investigated, with the aim to improve the practical value of blood glucose prediction.