Developing a hybrid linear model with a multilayer feed-forward neural network for HbA1c modeling among diabetes patients

Hemoglobin A1c (HbA1c) is the gold-standard measure for diagnosing and managing diabetes. Given importance of data-driven decisions, this paper aimed to develop a method elucidating predicting HbA1c levels. We developed comprehensive analyzing multiple linear regression through R syntax, embedding multilayer feedforward neural networks (MLFFNN) bootstrapping. The success proposed was determined by accuracy prediction. quality obtained model represented size minimum mean square error (MSE). This study used secondary diabetes data with total 1000 observations illustrate development (data after bootstrapping procedure). clinical relevance significance each preselected variable were evaluated before further testing. variables assessed using MLFFNN methodology, such as HbA1c, fasting blood sugar (FBS), urea, sodium It found that FBS, levels can all be verify HbA1c. FBS ( = 0.45931; Std SE= 0.01018; p< 0.01), urea =-0.03777; 0.00266; p < =-0.06685; 0.01112; 0.01) had significant impact on Our strategy provides an accurate prediction possible. methodology precisely assesses validity final model. Superior performance leads more efficient management in decision-making.