Penggunaan Kecerdasan Buatan untuk Menganalisis Faktor Risiko Diabetes dengan menggunakan Random Forest Classifier

Abstract

Diabetes is a non-communicable disease that deserves attention and poses a significant public health challenge. Although not a contagious disease, preventive measures and early detection of diabetes risk are crucial. This study used machine learning-based artificial intelligence to identify diabetes risk factors. The model was created using the Random Forest Classifier (RFC) algorithm, which has 16 variables as parameters. The model was built using the Python programming language, with data collection spanning from 2015 to 2018. The research included needs analysis, data collection, data preprocessing, model training, predictive model creation, system design, implementation, and testing. The final results showed that, with an accuracy of 89%, the model could be used effectively to predict diabetes risk. Furthermore, the model identified more pre-diabetes classes than other classes.