Performansi K-NN, J48, Naive Bayes dan Regresi Logistik sebagai Algoritma Pengklasifikasi Diabetes

Agung Mulyo Widodo (Universitas Esa Unggul)
Yanathifal Salsabila Anggraeni (Universitas Esa Unggul)
Nizirwan Anwar (Universitas Esa Unggul)
Arief Ichwani (Universitas Esa Unggul)
Binastya Anggara Sekti (Universitas Esa Unggul)

Prosiding Seminar Nasional Sisfotek (Sistem Informasi dan Teknologi Informasi)
Vol 5 No 1 (2021): SISFOTEK V 2021

Diabetes is a chronic disease characterized by high blood sugar (glucose) levels. This disease is often found in adults who are elderly, but this disease can also attack people who are still young. Along with advances in machine learning technology to support decision makers, many predictive models are made of whether a person can be classified as diabetic or not by using certain algorithms. In this study, a prediction model was made whether a person is classified as diabetic or not, based on parameters/variables, namely weight, height, cholesterol levels, fasting sugar, non-fasting sugar, uric acid levels and gender. Prediction model is made using K-NN, J48 (based on decision tree), Naive Bayes and logistic regression classification algorithms. Then a performance analysis was carried out on the testing results of each of these algorithms, and it was found that the K-NN algorithm produced a prediction model with the highest accuracy compared to the three algorithms used in this study.