Hi,

i am not sure about your understanding of "data pertubation", but if you are using Matlab, the ensemble methods typically provide a method named predictorImportance, which should quantify the importance of the features for/in the ensemble.

Greetings, Sebastian

Hi,

Do you mean permutations of your dataset?

Could you be more specific about your problem so that we can provide a better answer?

If you are dealing with regression/classification caret package in R is an excellent package.

If you are looking for inference based feature selection, i.e. using data permutation, check out the ranger package in R for variable selection using random forest.

Stefano

Hi Poornima,

The answer is Yes, WEKA can be used to do such task using the "Genetic" search algorithm.

HTH.

Samer Sarsam, PhD.

Hi,

I hope that the following work is what you are searching for:

https://hal.inria.fr/hal-01084301/document

Its implementation can be found here:

https://cran.r-project.org/web/packages/RFgroove/RFgroove.pdf

Best regards,

Anis.