About average impurity decrease from RandomForest

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

About average impurity decrease from RandomForest

lgarciaag89

Hi

I have a doubt respect the values of impurity decrease from RandomForest,
respect to a regression problem weka use a function based on varience for
compute the impurity decrease values. for example for the next command:

/*java weka.classifiers.trees.RandomForest -t data/iris.arff
-attribute-importance*/

a output part has this form:

/*Attribute importance based on average impurity decrease (and number of
nodes using that attribute)

      0.63 (   238)  petallength
      0.55 (    67)  sepalwidth
      0.55 (   133)  sepallength
      0.53 (   168)  petalwidth*/

if i want compare the attributes using the average impurity decrease, ¿Do I
have to minimize or maximize with respect to average impurity decrease?

¿where are reported the function based on variance for compute the impurity
decrease?

¿How i can get number of nodes using that attribute values?

best regards,
Luis



--
Sent from: https://weka.8497.n7.nabble.com/
_______________________________________________
Wekalist mailing list -- [hidden email]
Send posts to [hidden email]
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