Classification and diagnosis of diabetic with neural network algorithm learning vector quantizatin (LVQ)

Determining the type of Diabetes Mellitus (DM) is very important to determine what
treatment is suitable for a patient. Unfortunately patient information about what type of diabetes
is often ignored, so the patient gets a wrong diagnosis. This study aims to build a classification
model in determining a DM patient diagnosed with one type of DM, namely type 1 DM, type 2
DM, Gestational DM or special type DM. The indicators used in determining the classification
for diagnosing patients are age, sex, blood pressure, levels of blood glucose, weight, and height.
The classification method used is the Neural Network method with Learning Vector
Quantization (LVQ) algorithm. Algorithm LVQ provides results 96% accuracy for training data
with final epoch is 759 and 90% accuracy for testing data.