Skip to main content

Table 3 Values of the comparative statistics (after the 10-fold cross-validation approach) of the different algorithms tested in a preliminary analysis of species distribution and productivity models

From: Predicting current and future suitable habitat and productivity for Atlantic populations of maritime pine (Pinus pinaster Aiton) in Spain

Species distribution model
  ANN CTA FDA GAM GBM GLM MARS RF SRE
  AUC 0.7866 (0.0201) 0.7589 (0.0175) 0.7812 (0.0160) 0.7902 (0.0188) 0.7917 (0.0162) 0.782 (0.0176) 0.785 (0.0183) 0.807 (0.0206) 0.620 (0.0164)
Productivity model
  MLR kNN MARS RT RF     
  R2 0.4135 (0.2298) 0.4433 (0.2585) 0.5204 (0.1638) 0.4728 (0.1980) 0.5954 (0.2309)     
  1. Comparative statistics: area under the ROC curve (AUC) and pseudo-coefficient of determination (R2). Values in brackets are the standard deviations for the 10 predictions
  2. ANN artificial neural networks, CTA classification tree analysis, FDA flexible discriminant analysis, GAM generalized additive models, GBM generalized boosted regression models, GLM generalized linear models, MARS multivariate adaptive regression splines, RF random forest, SRE rectilinear surface range envelop (equivalent to BIOCLIM), MLR multiple linear regression, KNN nearest neighbour, regression trees using M5P algorithm