Additional file 1 of Detection of diabetic patients in people with normal fasting glucose using machine learning

Additional file 1: Fig. S1. Training and validation loss of DNN. Fig. S2. The other 11 characteristics with significant differences between diabetic and non-diabetic individuals with normal fasting glucose. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001. Fig. S3. Correlation of all features in the training set. Fig. S4. Feature importance ranking of the models constructed by mRMR-selected features. Fig. S5. Number of diabetic patients towards different thresholds of normal fasting glucose. Orange point is a turning point that the number of individuals with diabetes has halved when using 5.69 as the threshold of normal fasting glucose.