Perbandingan RFE dan SelectKbest untuk Klasifikasi Penyakit Diabetes dengan Random Forest

Abstract

Diabetes is a condition that happens in our metabolic system characterized by high level of blood sugar or known as hyperglycemia. Hyperglycemia can either be caused by auto immune insulin destruction problems or insulin resistance in the body. According to World Health Organization, nearly 350 million people suffers from diabetes. Several unwanted side effects can occur from diabetes such as blindness, amputation, and kidney failures if they aren’t aware of the disease. Sadly, not many people know the dangers of diabetes. Therefore, a machine that can accurately and efficiently classify diabetes from its symptoms is our top priorities. On this research SelectKBest feature selection when paired with Random Forest Algorithm is fairly accurate at classifying and predicting diabetes with accuracy and recall value of 0.72 each.