UJI KORELASI DAN ANALISIS CLUSTERING GULA DARAH PUASA, KOLESTEROL TOTAL, TRIGLISERIDA, SERTA ASAM URAT

Clustering Partitional Technique, Diabetes, K-Means Algorithm, Pearson Correlation

Abstract

Diabetes  is a disease that occurs due to increase blood sugar levels as a result of the body's ability to produce insulin hormones normally or the body cannot properly utilize the hormone insulin produced. This study aims to find the relationship between fasting blood sugar (FBS) to total cholesterol, triglyceride levels, and uric acid levels in patients with diabetes and classify the data using clustering partitional technique. This study, conducted by taking a sample of 32 patients with diabetes mellitus who performing laboratory tests for levels of FBS, total cholesterol, triglycerides, and uric acid. Data obtained from the medical record of RSUD R Soeprapto Cepu, Central Java in April 2014. The data of treated patients included FBS (mg/dL), total cholesterol (mg/dL), triglycerides (mg/dL), and uric acid (mg/dL). Correlation test using Pearson correlation where data clustering analysis using the K-Means algorithm. The result of Pearson correlation analysis using SPSS 16.0 tools is that there is no correlation between FBS levels and total cholesterol (p> 0.05, sig = 0.174) also no relationship between FBS levels and uric acid levels (p> 0.05, sig = 0.868). However, there is a relationship between FBS levels with triglycerides (p> 0.05, sig = 0.000). Clustering process classifies data into two clusters. Final cluster center fix after four times iteration with minimum distance 481.662.