Klasifikasi Menggunakan Metode Random Forest untuk Awal Deteksi Diabetes Melitus Tipe 2

Abstract

Type 2 diabetes mellitus (T2DM) is a chronic disease with increasing prevalence. Early detection of DMTP2 is crucial in managing and preventing this disease. In this study, we propose the use of Random Forest method for early classification of T2DM based on risk factors. The dataset was obtained from UPTD Puskesmas Jatiroto with a total of 1111 data with 6 attributes of DMTP2 factors and 1 label. In the pre-processing stage, initial data processing includes cleaning missing values, feature engineering, and separation of training and test data. Next, the Random Forest model is trained using data that has been validated using K-Fold Cross Validation. Experimental results show that the proposed model produces an average accuracy of each fold of 97%. The final stage of evaluating the model by calculating precission, recall, and F1-Score, respectively, obtained results of 95%, 97%, and 96%. Model evaluation focuses on predicted labels so that the model can predict well in the case of DMTP2 problems based on similar data configurations.