Hello.
I'm trying to use the ENMeval package to get the optimal RM and FC values in the Maxent model.
The Maxent model I'm trying to spin is trying to set the replicated run type to bootstrap. Because bootstrap is more appropriate than cross-validation because I have fewer coordinates.
In ENMeval, the optimal value can be determined by the AICc value obtained by setting several FC and RM values.
ENMeval's parts include randomkfold, block, and checkerboard methods.
However, there is no bootstrap verification in ENMeval.
Looking for other papers, ENMeval used the checkerboard method, obtained the optimal FC and RM values, and used differently, such as using subsample in the Maxent model.
1. Is it okay to have different verification methods like this?
2. Or can I take out parts from ENMeval and get FC and RM values, and then apply it to Maxent models?
3. Or I wonder if ENMeval partitions and Maxent's replicated run type are different.
The sentence may be awkward because I used the translator.......
+
Article ENMeval: An R package for conducting spatially independent e...
ENMeval github:
https://jamiemkass.github.io/ENMeval/articles/ENMeval-2.0-vignette.html
- Badges
- Science topic
- Similar topics
- Filing