Re: st: Classification analysis/risk tree (CART) in Stata - anything more?

[Ron�n Conroy <rconroy@rcsi.ie>]

On 21 Aib 2007, at 19:19, Clare L Maxwell wrote:

> Has Stata thought about this any more?  Especially, might CART be a  
> possibility in Stata within the next year?
The problem with this technique is that it is very very un-Stata-like.

I am currently carrying out a classification analysis to help us  
identify particular outcome groups within a large medical database.

The trouble is that you need to interact with the process to make  
sure that the tree is optimised for real life. For example, you might  
decide to use readily-available information or cheap tests for early  
splits and only to use more difficult-to-access information or more  
expensive tests if this offered the promise of identifying very  
useful high- or low-risk subsets. So there are times when you may  
want to force a split at a particular node, or to override the split  
value with one that makes for easier clinical practice, etc.

This is very straightforward with a point-and-click package like JMP  
where each node has a popup menu of things you can do and a popup  
list of split candidates for each node. But I cannot imagine how you  
would make this all work in Stata.





=========
Ron�n Conroy
Royal College of Surgeons in Ireland
rconroy@rcsi.ie
+353 (0) 1 402 2431
+353 (0) 87 799 97 95
http://www.flickr.com/photos/ronanconroy




*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/