Table 2.
Accuracy metrics for species distribution models of Dermacentor variabilis ticks built with five modeling methods including GLM, MaxEnt, GAM, RF, and BRT
Method | Dataset | AUC | TSS | Deviance | Mean omission |
---|---|---|---|---|---|
GLM | Positive | 0.92 | 0.76 | 0.23 | 0.15 |
All | 0.90 | 0.70 | 0.80 | 0.15 | |
GAM | Positive | 0.92 | 0.79 | 0.70 | 0.11 |
All | 0.95 | 0.79 | 0.55 | 0.12 | |
MaxEnt | Positive | 0.95 | 0.83 | 0.23 | 0.12 |
All | 0.95 | 0.77 | 0.66 | 0.12 | |
BRT | Positive | 0.91 | 0.77 | 0.27 | 0.13 |
All | 0.90 | 0.69 | 0.89 | 0.15 | |
RF | Positive | 0.93 | 0.81 | 0.20 | 0.10 |
All | 0.96 | 0.79 | 0.47 | 0.11 |