Application of the Gaussian Naive Bayes Algorithm for Diabetes Mellitus Risk Classification Based on Simulated Patient Medical Record Data

Penerapan Algoritma Gaussian Naive Bayes Untuk Klasifikasi Risiko Diabetes Melitus Berdasarkan Data Simulasi Rekam Medis Pasien

Abstract

Diabetes mellitus is a chronic disease with a number of cases that tends to increase over time and can lead to various complications if it is not recognized and treated at an early stage. In this context, the application of data mining techniques can be used to identify and classify the risk level of diabetes based on patients’ characteristics, thereby supporting a more effective decisionmaking process. This study applies the Gaussian Naive Bayes algorithm to classify the risk of diabetes mellitus using 200 simulated data records representing the characteristics of patients medical records. The data consist of several attributes, including age, gender, body weight, height, Body Mass Index (BMI), blood pressure, glucose level, cholesterol level, physical activity, family history of diabetes, smoking habits, and diabetes status as the target variable. The research process consists of several stages, beginning with Exploratory Data Analysis (EDA) to understand the characteristics of the dataset, followed by data preprocessing, conversion of categorical data using the Label Encoding method, and division of the dataset into training and testing sets using an 80:20 ratio. Subsequently, a classification model was developed using the Gaussian Naive Bayes algorithm, and its performance was evaluated using several evaluation metrics, including Accuracy, Precision, Recall, F1-Score, Confusion Matrix, and Classification Report. Based on the testing results, the model achieved an Accuracy of 97.50%, Precision of 94.74%, Recall of 100%, and F1-Score of 97.30%. These results indicate that the Gaussian Naive Bayes algorithm is capable of providing highly accurate classification results on the simulated dataset used in this study. Therefore, the algorithm has the potential to serve as a supporting approach for the identification and classification of diabetes mellitus risk.