Diabetes Prediction Using Machine Learning

IEEESTEC 17TH (2024), (pp. 63–66)

 

АУТОР / AUTHOR(S): Jelena Nedeljković

Download Full Pdf    

DOI: 10.46793/IEEESTEC17.063N

САЖЕТАК / ABSTRACT:

In this paper, we present the application of machine learning, specifically the Support Vector Machine (SVM) algorithm, for diabetes prediction using the Pima Indians Diabetes dataset. Diabetes is a serious global health issue, and timely diagnosis can significantly improve treatment outcomes. The PIMA dataset contains 768 instances with 8 medical attributes, including glucose concentration, body mass index (BMI), blood pressure, and family history of diabetes. The process involves data preprocessing, which includes cleaning, standardization, and splitting the data into training and testing sets. We used the SVM algorithm for classification, and the model was evaluated using accuracy metrics. Our model achieved an accuracy of 78.6% on the training data and 77% on the test data, which is considered a solid result given the size and characteristics of the dataset. The results show that SVM is effective for diabetes prediction, although a larger dataset could further improve model performance. In future work, we plan to test other algorithms, such as Random Forest and Neural Networks, to achieve even higher prediction accuracy.

КЉУЧНЕ РЕЧИ / KEYWORDS:

Machine learning, SVM, diabetes

ЛИТЕРАТУРА/ REFERENCES: