Method | Az | σ | SE |
---|
LR | 0.781 | 0.029 | 0.009 |
LRint | 0.798 | 0.029 | 0.009 |
k | 0.947 | 0.015 | 0.004 |
kx1 | 0.852 | 0.018 | 0.005 |
kx2 | 0.734 | 0.023 | 0.008 |
- This table gives the area under the receiver operator characteristic curve (Az) quantities for the standard logistic regression model using x1 and x2 (LR), the logistic regression model using x1 and x2 with x1 × x2 interaction (LRint), the statistical learning (SL) model using a kernel mapping with x1 and x2 simultaneously (k), and partial SL-kernel models using x1 (kx1) and x2 (k2) individually. Az, and σ are the respective means and standard deviations derived from processing the 10 validation datasets with the trained models. SE is the standard error in Az