Blood Glucose Prediction Using Convolutional Long Short-Term Memory Algorithms

Diabetes Mellitus is one of the preeminent causes death to date. Effective procedures are necessary prevent diabetes and avoid complications that may cause early death. A common approach control patient blood glucose, which necessitates a periodic measurement glucose concentration. This study developed prediction system using convolutional long short-term memory (Conv-LSTM) algorithm. Conv-LSTM variation LSTM algorithms suitable for use in time series problems. overcomes lack algorithm because latter cannot access content previous cells when its output gate has closed. We tested varied experiment check effect cross-validation ratio between 70:30 80:20. The indicates data split more stable compared with 80:20 split. best result shows measure 21.44 RMSE 8.73 MAE. With application conv-LSTM correct parameters selected split, our attains accuracy comparable regular LSTM.