[R] Tree vs. Rpart

From: Rizkalla, Carol Elisabeth <crizkall_at_purdue.edu>
Date: Sat 11 Feb 2006 - 07:44:33 EST

I'm using classification trees for the first time.

I understand the difference between these 2 packages, but I'm having a bit of trouble interpreting the results.

I have 3 different response variables, but I'll only use 1 in this discussion.  

I first ran Tree. I was happy with the results, 6 nodes, everything made sense. Misclassification rate of 15%.

Then I ran cross-validation and it showed the optimal tree was only 2 nodes.  

I then ran Rpart, it provides the optimal tree with 2 nodes. But of course, this doesn't provide much explanation for the response.

Additionally, the misclassification rate increased to 24%.  

Is it correct to say that if I want to only describe my results, I can use the Tree result.

But if I want a predictive model, I should use the Rpart results, even though it had a higher misclassification?  

Thank you,


Carol Rizkalla

Graduate Research Assistant

195 Marsteller St.

Purdue University

West Lafayette, IN 47907

(765) 494-3997  

Sentiment without action is the ruin of the soul. - Edward Abbey  

        [[alternative HTML version deleted]]

R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Sat Feb 11 07:49:23 2006

This archive was generated by hypermail 2.1.8 : Mon 13 Feb 2006 - 22:42:10 EST