Springer Nature
Browse
12902_2019_436_MOESM1_ESM.docx (25.27 kB)

MOESM1 of Predictive models for diabetes mellitus using machine learning techniques

Download (25.27 kB)
journal contribution
posted on 2019-10-16, 04:01 authored by Hang Lai, Huaxiong Huang, Karim Keshavjee, Aziz Guergachi, Xin Gao
Additional file 1: Table S1. Summary of characteristics of patients in the dataset. Table S2. Confusion Matrix for the Gradient Boosting Machine (GBM) model with the threshold of 0.24. Table S3. Confusion Matrix for the Logistic Regression model with the threshold of 0.24. Table S4. Confusion Matrix for the Random Forest model with the threshold of 0.24. Table S5. Confusion Matrix for the Rpart model with the threshold of 0.18. Table S6. Comparing the AROC with other machine-learning techniques using the class weight method. Table S7. Sensitivity, Specificity, Misclassification Rate, and AROC values of the four models on the studied data set. Table S8. Sensitivity, Specificity, Misclassification Rate, and AROC values of the four models on the PIMA Indians data set.

History