Komparasi Metode Klasifikasi terhadap Data Penderita Penyakit Diabetes Menggunakan Python 3

Abstract

Diabetes is a serious challenge in the world of health, with broad impacts. In an effort to overcome this problem, it is important to analyze the classification of diabetes data to provide valuable insights. This study focuses on the comparison of the two main classification methods, namely Naive Bayes and Support Vector Machine (SVM), in analyzing diabetes data. We use the Python 3 programming language for implementation. The initial study involved the characterization of the dataset, including parameters such as blood pressure and blood glucose levels, which were important factors in the analysis. The preprocessing process is carried out to ensure data quality by overcoming missing or invalid values. After that, the dataset is divided into training and testing subsets. The Naive Bayes and SVM methods are implemented using the scikit-learn library in Python 3. Both models are trained using a training subset and tested on a test subset. The test results show that both methods have good performance in classifying diabetes data, but SVM stands out with higher accuracy. SVM has the ability to handle complex data and find optimal decision boundaries. The Naive Bayes model achieves the highest accuracy of 78.13% on 70% training data and 30% testing data, while the SVM model achieves 79.63% on 90% training data and 10% testing data. Overall, this study provides an in-depth understanding of the effectiveness of both methods in the context of classifying data on diabetics.