SIHALOHO, JUAN ANDREAS (2025) COMPARISON OF PREDICTION ALGORITHMS FOR PREDICTING DIABETES USING ORANGE DATA MINING. S1 thesis, UNIVERSITAS KATOLIK SOEGIJAPRANATA.

Preview	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_COVER.pdf Download (472kB) | Preview
	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_BAB I.pdf Restricted to Registered users only Download (482kB)
	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_BAB II.pdf Restricted to Registered users only Download (490kB)
	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_BAB III.pdf Restricted to Registered users only Download (800kB)
	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_BAB IV.pdf Restricted to Registered users only Download (823kB)
	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_BAB V.pdf Restricted to Registered users only Download (406kB)
Preview	Text 18.K1.0074_JUAN ANDREAS SIHALOHO_DAFPUS.pdf Download (472kB) | Preview
	Text 18.K1.0012_KEVIN HENDRAWAN S_LAMP.pdf Restricted to Registered users only Download (596kB)

Abstract

Diabetes is one of the diseases with the highest mortality rate and has experienced a very significant increase in the number of sufferers. According to the IDF Diabetes Atlas (2021) from International Diabetes Federation, 10.5% of the adult population (20-79 years) has diabetes, with almost half unaware that they are living with the condition. With machine learning technology, we can create models that can later predict whether someone has diabetes or not so they can treat it immediately. By using simplified data from The Behavioral Risk Factor Surveillance System (BRFSS) through surveys and uploaded to Kaggle, we can train models with various algorithms which can then predict whether someone has diabetes or not. There are 4 algorithms that are used in this study, namely Naive Bayes, Logistic Regression, Gradient Boosting, and Neural Network. The results of this study show that these algorithms can predict whether someone has diabetes or not with an accuracy of up to 74%.

Actions (login required)

View Item