Abstract:
Nowadays, Machine learning techniques play an increasingly prominent role in medical diagnosis because using these techniques can be analyzed to find patterns or facts that are difficult to explain, which contributes to making the diagnosis more accurate. The purpose of this research is to compare the efficiency of diabetic classification models with and without interaction using four machine learning techniques including Decision tree, Random forest, Support Vector Machine and K-Nearest neighbor. These models are compared base on accuracy, precision, recall, and F1-score. The results of this research showed that the models with interaction have better classification performance than those without interaction for all 4 machine learning techniques. Among models with interaction, Random forest classifiers had the best performance with 97.5% accuracy, 97.4% precision, 96.6% recall, and 97% F1-score. In the same way, Random forest also had the best classification performance among models without interaction with 88.2% accuracy, 92.2% precision, 89.3% recall, and 90.7% F1-score. The findings from this research can be further developed into a program to effectively screen diabetes patients.