Table 3.
Sample | Best Model | Cross Validation | Prediction Error | p value |
Whole | 2 (Age, Gender) | 10 | 33.1 % | < 0.001 |
Non-obese group | 3 (Age, ADRB2-16, UCP3) | 10 | 32.5 % | < 0.001 |
Obese group | - | - | - | n. s. |
Sample: sample of the analysis. Best Model: number of factors and which factors best predict the disease status in sample. Cross Validation: value of the consistency of the model across the 10 repeated test. Prediction Error: percent of subjects incorrectly classified following the proposed model. p value: significant level calculated through a 1000-fold permutation method.