TABLE 4.
Akaike information criterion (AIC) and Schwarz criterion (SC) values of the selected, full model (model with all available predictors), and null model (model with coefficient of predictors as 0) trained by ordinal and binary logistic regressions using color and vegetation indices from proximal and aerial RGB images.
| Ordinal logistic | Binary logistic | ||||
|
|
|
||||
| Model | Proximal | Aerial | Proximal | Aerial | |
| Selected | AIC | 286.6 | 201.0 | 175.7 | 110.5 |
| SC | 305.3 | 234.3 | 185.0 | 126.1 | |
| Full | AIC | 291.1 | 208.9 | 184.3 | 104.4 |
| SC | 344.8 | 253.2 | 221.7 | 141.9 | |
| Null | AIC | 376.4 | 344.2 | 232.2 | 209.9 |
| SC | 385.7 | 358.1 | 235.3 | 213.0 | |
A low value of AIC and SC signifies good statistical fit of the model.