ANALISA KESALAHAN PENGINPUTAN DATA MENGGUNAKAN METODE KNN UNTUK DATA PENYAKIT DIABETES

Missing value in data is still a problem in data analysis in various research, especially in the business sector. These problems will certainly influence decisions in determining wrong or inappropriate business decisions. Data Preprocessing phase, Simulation of Missing Value phase, Modeling phase, Performance Evaluation phase. In this study the K-Nearest Neighbor (K-NN) method was used to input missing data.
Diabetes is a disease characterized by glycemic disturbances, including chronic sustained hyperglycemia and acute glucose fluctuations. Because diabetes is closely related to the body's metabolism, observing blood vessels is very important. Observations were made using Mean Amplitude of Glycemic Excursion (MAGE). Definitely, MAGE is an important variable to solve the clinical problem of diabetes which contributes to causing oxidative stress related to macro and microvascular complications. MAGE is technically used with continuous blood glucose data obtained via Continuous Glucose Monitoring (CGM). Because CGM cannot be used in everyday observations. The contribution of this research is the use of discrete data (3 day observations) for use in MAGE measurements.