Table 6. Baseline logistic regression model, generalized regression Elastic Net model (with AICc and Leave-One-Out-Validation) on the predictors of breast cancer from gene environment interactions including age as a factor.
| Logistic Regression Original Model with Validation | Generalized Regression Elastic Net Model | |||||
|---|---|---|---|---|---|---|
| With AICc Validation | With Leave-One-Out Validation | |||||
| Parameters | Estimate | p (X2) | Estimate | p (X2) | Estimate | p (X2) |
| (Intercept) | 1.7735 | 0.2386 | 1.2899 | 0.1972 | 1.9324 | 0.0027 |
| Age | -2.8420 | 0.0001 | -2.2898 | <0.0001 | -2.5734 | <0.0001 |
| BMI*Alcohol | 2.5349 | 0.1595 | 1.4491 | 0.2790 | 2.1891 | 0.0152 |
| Alcohol | -2.2623 | 0.1675 | -1.0589 | 0.3814 | -1.8443 | 0.0461 |
| BMI*MTR 2756 | -2.3623 | 0.2526 | -0.8735 | 0.1162 | -1.1353 | 0.0581 |
| MTR 2756 | 1.3848 | 0.4603 | 0 | 1.0000 | 0 | 1.0000 |
| BMI | 0.1668 | 0.9215 | 0.2466 | 0.8302 | 0 | 1.0000 |
| Misclassification Rate | 0.3000 | 0.2375 | 0.2278 | |||
| AICc | 50.10 | 94.79 | n/a | |||
| Area under the curve | 0.7656 | 0.8313 | 0.8455 | |||
Note. AICc: Akaike’s information criterion with corrections.