Safitri, Rizki Alifia (2024) Perbandingan Metode K-Nearest Neighbor (Knn) Dan Naïve Bayes: Klasifikasi Resiko Diabetes (Studi Kasus Posbindu Desa Bulupitu Kabupaten Malang). Undergraduate thesis, Fakultas Teknologi Informasi Universitas Merdeka Malang.

Preview	Text HALAMAN AWAL.pdf Download (1MB) | Preview
Preview	Text BAB I.pdf Download (240kB) | Preview
	Text BAB II.pdf Restricted to Repository staff only Download (282kB)
	Text BAB III.pdf Restricted to Repository staff only Download (390kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (399kB)
	Text BAB V.pdf Restricted to Repository staff only Download (265kB)
Preview	Text DAFTAR PUSTAKA.pdf Download (233kB) | Preview
Preview	Text HASIL CEK PLAGIASI.pdf Download (343kB) | Preview

Abstract

Diabetes mellitus is among the fastest-growing health problems in the 21st century. One of the causes is the lack of public awareness for regular health check-ups, while the lifestyle they lead is generally unhealthy. The hemoglobin A1c (HbA1c) test is highly recommended for detecting diabetes, but this service is not yet available at the Posbindu in Bulupitu Village. Therefore, an alternative approach is needed to detect the risk of diabetes early, using data mining. The data mining methods used in this study are the Naïve Bayes and k-Nearest Neighbor (kNN) classification methods. The variables used to determine diabetes risk include gender, age, family history of diabetes, frequent urination, Body Mass Index (BMI), blood sugar level, and the output of diabetes risk. The dataset was split for testing and training using cross-validation and various ratios (60:40, 70:30, 80:20, and 90:10). The best accuracy for the Naïve Bayes method was obtained with a k-fold cross-validation split of 2, yielding 96.1%. Meanwhile, the best results for the kNN method were obtained with an 80:20 dataset ratio, with the Manhattan distance being the most effective distance calculation compared to Euclidean distance and Chebyshev distance

Actions (login required)

View Item