Pages: [1]
  Print  
Author Topic: EqualLabelWeighting  (Read 975 times)
csummerhill
Newbie
*
Posts: 3


« on: February 10, 2009, 07:52:34 PM »

New to this so I appreciate your patience...

I found on the forum that in order to compensate for an unbalanced class/skewed class distribution one should use the 'EqualLabelWeighting' operator.  So I loaded the data using a csv file and the 'ExampleSource' operator making the skewed column the label with the attribute wizard, added the 'EqualLabelWeighting' operator and hit play...It appears to have added a weight column and the weights appear to be appropriate to balance the column...what next.

I wanted to use a learner (like a DecisionTree) next but now the label is assigned to the wrong column...how do I change the label back to the variable I'm trying to predict for?  Is this even the right way to use the operator?

I'm lost...any help would be appreciated. 
Logged
csummerhill
Newbie
*
Posts: 3


« Reply #1 on: February 11, 2009, 05:39:29 AM »

found the 'ChangeAttributeRole' operator...this solved my issue...
Logged
Pages: [1]
  Print  
 
Jump to: