10 percentile logistic training threshold for future projections

[Ben Reed]

Hi all, I'm running MaxEnt to project changes in the distributions of Australian parrot species due to climate change. I have run MaxEnt with current environmental variables/occurrence data and used a projection of future climate change in said variables. For my final output I'm attempting to generate two maps; one of current distributions and the other of future distributions, using the 10th percentile training presence logistic threshold as a threshold for presence/absence. 

This works great for the current scenario (using 10th percentile training presence logistic threshold from MaxentResults.csv), however in MaxentResults.csv I honestly have no idea which output I should use to apply a similar threshold to the future projection, as the 10th percentile training presence logistic threshold lies outside of the dataspread of the future output (I am assuming it is the threshold for the present scenario). 

Are there any outputs from MaxentResults.csv I might use to apply a similarly appropriate threshold to my future projected scenario? 

Thanks for any help.

Ben

[swammy50]

The threshold you get from the current model is what you use for the future projections also. So if 0.25 is the value of the threshold you use this value for current and future. 

Sam

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at http://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

[Madan K. Suwal]

Hi, 

I also got the same problem. I tried with two thresholds, 10 % of presence and minimum of presence, 

in both the cases, the future prediction and past prediction value spread is lower than the threshold value. 

how to tackle this issue, 

thanks in advance.