Skip to main content

Table 2 Evaluation parameters for species distribution models built using human population data

From: Modeling the distribution of the West Nile and Rift Valley Fever vector Culex pipiens in arid and semi-arid regions of the Middle East and North Africa

Model

Parameters in final model

Observed training AUC

Observed test AUC

Null model training AUC

Null model test AUC

Prevalence

Entropy

AICc

Omission rate3N = 79

Test COR (N = 158)

Test deviance (N = 158)

Maxent Pop1

9

0.938

0.872

0.827 ± 0.01

0.450 +/- 0.03

0.151

8.043

8383

0.206

0.470

197.9

Maxent NoPop2

14

0.916

0.879

0.848 ± 0.01

0.476 +/- 0.04

0.185

8.252

8916

0.194

0.534

178.3

  1. 1Data sources used for model development: Bioclim + Vegetation + Population (N = 24).
  2. 2Data sources used for model development: Bioclim + Vegetation (N = 23).
  3. 3Using equal test sensitivity and specificity threshold. Maxent Pop threshold =0.22, Maxent No Pop threshold =0.34.
  4. Evaluation parameters for Culex pipiens distribution models generated using a set of environmental predictors including human population density (“Maxent Pop”) and a set excluding human population density (“Maxent No Pop”). Models were built using 322 presence points. “Null Model” parameters represent the average value of models built from one hundred 322-point dummy data sets. “Test AUC” and “Omission Rate” are calculated from an independent data set of 79 presence points, not used in training the models. “Test COR” and “Test Deviance” are calculated from a data set that includes the 79 test data points and 79 background points. “Test COR” = Pearson correlation coefficient.