Analyzing the amount of hemoglobin A1c using Singular Spectrum Analysis

Monthly hemoglobin A1c levels are a proper estimator in patients with diabetes type II for adequate diabetes control. In this paper, we apply Singular Spectrum Analysis (SSA), which is a non-parametric powerful method for forecasting and time series analysis, to forecast the amount of hemoglobin A1c. The results of two forecasting approaches in the SSA, namely vector and recurrent forecasting, are compared to those from other techniques including ARIMA, ARFIMA, BATS, TBATS, and ETS. The comparison is made using the two accuracy measures Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE).