Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2023
- Goals
- Record
- 1e9f0a8d-11c3-4e20-933c-598436a55b1e
- Published
- 2026-10-30
Universitas Dian Nuswantoro conducted an activity on the performance analysis of machine learning algorithms for diabetes diagnosis prediction. The activity involved Tominanto, Tominanto, Syukur, Abdul, Affandy, and Marjuni, Aris as participants. It addressed diabetes as a chronic condition characterized by blood sugar levels increasing above average. Over the long term, diabetes can have damaging effects on the kidneys, heart, nerves, eyes, and may cause other complications.
The activity focused on early diabetes prediction as an approach that can prevent and reduce the risk of further disease complications and lead to improved treatment. The research used the Pima Indians Diabetes Dataset, sourced from the UCI machine learning repository. Seven machine learning classification algorithms were applied: Support Vector Machine, Naive Bayes, K-Nearest Neighbor, Logistic Regression, Random Forest, Decision Tree, and Gradient Boosting.
The performance of the algorithms was evaluated using Accuracy, Precision, F-Measure, Recall, and Receiver Operating Characteristic. The results showed that K-Nearest Neighbor obtained the highest Accuracy at 80.51% and Precision at 78%. Random Forest achieved the highest Recall at 75%, F-Measure at 72%, and Receiver Operating Characteristic at 78.18%. These findings provide a comparison of the evaluated algorithms for diabetes diagnosis prediction. The activity is institutionally framed as research addressing early diabetes prediction and improved treatment through the evaluation of machine learning classification algorithms.