Application of Machine Learning Models for Early Detection and Accurate Classification of Type 2 Diabetes

Aug 2, 2026

Upload by Ivy

Early detection of diabetes is essential to prevent severe complications and improve patient outcomes. This work develops machine learning approaches to detect and classify type 2 diabetes while selecting the most effective model for predicting risk. Five models—K-nearest neighbor, Bernoulli Naïve Bayes, decision tree, logistic regression, and support vector machine—are evaluated using a Kaggle Pima Indian dataset of 768 patients with and without diabetes and multiple clinical variables. Results indicate K-NN achieves the highest accuracy at 79.6%, outperforming BNB at 77.2% and supporting ML as a promising diagnostic aid.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages16

Reading Time40 min

AccessFree to download

Abstract

Introduction

Diabetes background and risks

Type 1 versus type 2 diabetes

Methods

Machine learning models evaluated

Dataset and input variables

Results

Model performance comparison

Best-performing model accuracy

Discussion

Implications for early detection

diagnostics Article Application of Machine Learning Models for Early Detection and Accurate Classiﬁcation of Type 2 Diabetes Orlando Iparraguirre-Villanueva 1, Karina Espinola-Linares 2, Rosalynn Ornella Flores Castañeda 3 and Michael Cabanillas-Carbonell 4, * Citation: Iparraguirre-Villanueva, O.; Espinola-Linares, K.; Flores Castañeda, R.O.; Cabanillas-Carbonell, M. Application of Machine Learning Models for Early Detection and Accurate Classiﬁcation of Type 2 Diabetes. Diagnostics 2023, 13, 2383 . [https://doi.org/10.3390/](https://doi.org/10.3390/)[ ](https://doi.org/10.3390/)diagnostics13142383 Academic Editor: Marijn Speeckaert Received: 31 May 2023 Revised: 23 June 2023 Accepted: 7 July 2023 Published: 15 July 2023 Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . 1 Facultad de Ingenier½a y Arquitectura, Universidad Autânoma del Peró, Lima 15842, Peru; [oiparraguirre@ieee.org](oiparraguirre@ieee.org) 2 Facultad de Ingenier½a, Universidad Tecnolâgica del Peró, Chimbote 02710, Peru; [c24051@utp.edu.pe](c24051@utp.edu.pe) 3 Facultad de Arquitectura e Ingenier½a, Universidad C²sar Vallejo, Lima 15314, Peru; rﬂ[oresc@ucv.edu.pe](oresc@ucv.edu.pe) 4 Facultad de Ingenier½a, Universidad Privada del Norte, Lima 15083, Peru * [Correspondence: mcabanillas@ieee.org](Correspondence: mcabanillas@ieee.org) Abstract: Early detection of diabetes is essential to prevent serious complications in patients. The purpose of this work is to detect and classify type 2 diabetes in patients using machine learning (ML) models, and to select the most optimal model to predict the risk of diabetes. In this paper,ﬁve ML models, including K-nearest neighbor (K-NN), Bernoulli Naïve Bayes (BNB), decision tree (DT), logistic regression (LR), and support vector machine (SVM), are investigated to predict diabetic patients. A Kaggle-hosted Pima Indian dataset containing 768 patients with and without diabetes was used, including variables such as number of pregnancies the patient has had, blood glucose concentration, diastolic blood pressure, skinfold thickness, body insulin levels, body mass index (BMI), genetic background, diabetes in the family tree, age, and outcome (with/without diabetes) . The results show that the K-NN and BNB models outperform the other models. The K-NN model obtained the best accuracy in detecting diabetes, with 79.6% accuracy, while the BNB model obtained 77.2% accuracy in detecting diabetes. Finally, it can be stated that the use of ML models for the early detection of diabetes is very promising. Keywords: diabetes; machine learning; classiﬁcation; modeling; analysis 1. Introduction Diabetes occurs when blood sugar levels rise due to metabolic problems. This typeof exposure can damage various organs and body systems, such as the heart, blood vessels, and eyes. It is important to note that these adverse effects are directly caused by hyperglycemia, which is elevated blood sugar levels. This is because the body has trouble controlling blood sugar levels or cannot properly use the insulin it produces [1] . The hormone insulin helps glucose reach and be available to the cells. Note that diabetes is divided into two main categories, type 1 and type 2 . To fully understand type 1 diabetes, it is important to know that it is an autoimmune disease, which means that the body's immune system constantly attacks and destroys insulin-producing cells. Type 2 diabetes is characterized by problems with the proper use of insulin produced by the body due to factors related to an individual's lifestyle [2] . During the last decade, asigniﬁcant increase in the prevalence of type 2 diabetes has been observed in

FAQs

What is the goal of the study on type 2 diabetes?

The study aims to detect and classify type 2 diabetes using machine learning models and to identify the most optimal model for predicting diabetes risk.

Which machine learning models are compared in the work?

Five models are evaluated: K-nearest neighbor (K-NN), Bernoulli Naïve Bayes (BNB), decision tree (DT), logistic regression (LR), and support vector machine (SVM).

What dataset and variables are used for prediction?

The work uses a Kaggle-hosted Pima Indian dataset of 768 patients and includes features such as number of pregnancies, blood glucose concentration, diastolic blood pressure, skinfold thickness, body insulin, BMI, genetic and family diabetes background, age, and the diabetes outcome.

Aug 2, 2026

Upload by Ivy