Skip to contents

Evaluate a model

Source: vignettes/articles/evaluate-model.Rmd
evaluate-model.Rmd

Intro

In the previous articles you have learned how to prepare the data for the analysis, how to train a model and how to make predictions using SDMtune. In this article you will learn how to evaluate your model using three different metrics.

SDMtune implements three evaluation metrics:

  • AUC: Area Under the ROC curve (Fielding and Bell 1997)
  • TSS: True Skill Statistic (Allouche, Tsoar, and Kadmon 2006)
  • AICc: Akaike Information Criterion corrected for small sample size (Warren and Seifert 2011)

We will compute the value of the metrics on the training dataset, using the default model that we trained in a previous article.

AUC

As usually we first load the SDMtune package:

library(SDMtune)
#> 
#>    _____  ____   __  ___ __
#>   / ___/ / __ \ /  |/  // /_ __  __ ____   ___
#>   \__ \ / / / // /|_/ // __// / / // __ \ / _ \
#>  ___/ // /_/ // /  / // /_ / /_/ // / / //  __/
#> /____//_____//_/  /_/ \__/ \__,_//_/ /_/ \___/  version 1.3.2
#> 
#> To cite this package in publications type: citation("SDMtune").

The AUC can be calculated using the function auc():

auc(default_model)
#> [1] 0.8728322

We can also plot the ROC curve using the function plotROC():

plotROC(default_model)

TSS

The TSS is computed with the function tss():

tss(default_model)
#> [1] 0.6419

AICc

For the AICc we use the function aicc(). In this case we need to pass to the env argument the ‘predictors’ raster object that we created in the first article:

aicc(default_model, 
     env = predictors)
#> [1] 6618.725

Try yourself

Try to compute the three metrics using a model trained using the Maxnet method.

Conclusion

In this article you have learned:

  • how to calculate the AUC;
  • how to plot the ROC curve;
  • how to calculate the TSS;
  • how to calculate the AICc.

In the next article you will learn two different strategies that can be used to correctly evaluate the model performance.

References

Allouche, Omri, Asaf Tsoar, and Ronen Kadmon. 2006. “Assessing the Accuracy of Species Distribution Models: Prevalence, Kappa and the True Skill Statistic (TSS).” Journal of Applied Ecology 43 (6): 1223–32. https://doi.org/https://doi.org/10.1111/j.1365-2664.2006.01214.x.
Fielding, Alan H., and John F. Bell. 1997. “A review of methods for the assessment of prediction errors in conservation presence/absence models.” Environmental Conservation 24 (1): 38–49.
Warren, Dan L., and Stephanie N. Seifert. 2011. “Ecological Niche Modeling in Maxent: The Importance of Model Complexity and the Performance of Model Selection Criteria.” Ecological Applications 21 (2): 335–42. https://doi.org/https://doi.org/10.1890/10-1171.1.