s-news
[Top] [All Lists]

Variable importance for a tree analysis

To: s-news@lists.biostat.wustl.edu
Subject: Variable importance for a tree analysis
From: "Data Analytics Corp." <dataanalytics@earthlink.net>
Date: Tue, 16 Oct 2007 12:09:08 -0400
Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=dk20050327; d=earthlink.net; b=Pr/jeP8iyQ/DshfHVwgGPMZsN9tJqqqFQCCBrxjZZlxyGfGd/rksAKDi/n3OGx2R; h=Received:Message-ID:Date:From:User-Agent:MIME-Version:To:Subject:Content-Type:Content-Transfer-Encoding:X-ELNK-Trace:X-Originating-IP;
User-agent: Thunderbird 2.0.0.6 (Windows/20070728)
Good morning,

I'm using the S+8 tree function to do a CART analysis on a categorical variable. I have the tree I want after pruning plus graphs illustrating the tree and other results. I'd like to say something about which explanatory variables are the most important. My understanding is that this can be done, but a function is needed to rank the variables. I'd write one myself but I'm not sure at this time what's involved, or if "tree" provides the required information. Does anyone have a function to share or know where I can find one?

Thanks,

Walt Paczkowski

<Prev in Thread] Current Thread [Next in Thread>
  • Variable importance for a tree analysis, Data Analytics Corp. <=