Getting Feature Importance with RF in Experimenter

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Getting Feature Importance with RF in Experimenter

domhnallcarlin
Hi

Is it possible to generate (and output/save) the feature importances from Random Forest across multiple runs of 10f CV, while using the Experimenter via the gui? I can get them via Explorer, but when I enable the same option on the Experimenter, they don't appear anywhere in either the CSV or ARFF.

Thanks
Domhnall

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: Getting Feature Importance with RF in Experimenter

Eibe Frank-2
Administrator
In theory, it should be possible to get the full textual output of each model by switching the Experimenter into the "Advanced" mode and setting the "rawOutput" field in "CrossValidationResultProducer" to value "True".

Cheers,
EIbe

On Thu, Nov 21, 2019 at 11:45 PM Domhnall Carlin <[hidden email]> wrote:
Hi

Is it possible to generate (and output/save) the feature importances from Random Forest across multiple runs of 10f CV, while using the Experimenter via the gui? I can get them via Explorer, but when I enable the same option on the Experimenter, they don't appear anywhere in either the CSV or ARFF.

Thanks
Domhnall
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html

_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html
Reply | Threaded
Open this post in threaded view
|

Re: Getting Feature Importance with RF in Experimenter

dcarlin
This works, with the output being stored in the zip file defined in the -O
parameter of the CrossValidationResultsProducer (default
SplitEvaluatorOut.zip under weka-[version]/weka/).
Thanks!



--
Sent from: https://weka.8497.n7.nabble.com/
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to: To unsubscribe send an email to [hidden email]
To subscribe, unsubscribe, etc., visit
https://list.waikato.ac.nz/postorius/lists/wekalist.list.waikato.ac.nz
List etiquette: http://www.cs.waikato.ac.nz/~ml/weka/mailinglist_etiquette.html