Research on blood glucose data prediction based on I-GWO-KELM algorithm

Predicting future blood glucose (BG) levels in diabetic patients can help reduce high and low blood glucose events. The blood glucose prediction method in this article is as follows: the improved grey wolf optimizer is combined with the kernel extreme learning machine, called as I-GWO-KELM. The data set of patients was obtained by continuous glucose monitoring. Initially, the Kalman filter method was used to reduce the errors during data collection. The grey wolf optimizer algorithm optimizes the parameters of the kernel extreme learning machine, including regularization coefficients and kernel parameters. Then KELM was used to predict blood glucose, and mean absolute error (MAE) and root mean square error were (RMSE) used to evaluate the model performance. Experiments prove that BG prediction model based on I3-GWO-KELM outperformed other models, with MAE and RMSE are 12.127 mg/dL and 16.544 mg/dL for prediction horizon (PH) 30 minutes. I3-GWO-KELM has strong robustness and is suitable for the prediction of time series data.