R help - Jun 2012 - Good Decision Trees with Product Purchased Data?

A client has inquired about producing a decision tree from data which could
include:

- ID of brand purchased
- Importance ratings (1-10 scale) for a number of relevant attributes
(price, strength, recommended by a friend, etc.)  In other words, a rating
of how important each attribute is in the decision as to which brand to
purchase.

I've just run a test decision tree using the closest thing to a similar data
set, that I have at hand .  But only one attribute was selected for
plotting.  I used the "Tree" package for this test.

Question:
- Does one usually get good decision trees using data of this kind?

Thanks very much in advance to all for any info.

--
View this message in context:
http://r.789695.n4.nabble.com/Good-Decision-Trees-with-Product-Purchased-Data-tp4632438.html
Sent from the R help mailing list archive at Nabble.com.

Post on a statistics or data mining discussion site (e.g.
stats.stackexchange.com). There are tons of different algorithms for
fitting decision trees, and you are more likely to get an informative
discussion on their relative strengths and weaknesses there than here.
This is really NOT an R question.

-- Bert

Also, see below for a comment.

On Tue, Jun 5, 2012 at 1:42 PM, VikR <vikr at mindspring.com>
wrote:
> A client has inquired about producing a decision tree from data which could
> include:
>
> - ID of brand purchased
> - Importance ratings (1-10 scale) for a number of relevant attributes
> (price, strength, recommended by a friend, etc.) ?In other words, a rating
> of how important each attribute is in the decision as to which brand to
> purchase.
>
> I've just run a test decision tree using the closest thing to a similar
data
> set, that I have at hand . ?But only one attribute was selected for
> plotting. ?I used the "Tree" package for this test.
>
> Question:
> - Does one usually get good decision trees using data of this kind?
???
Define: "one" (who?)
Define: "usually"
Define: "good"
Define: "data of this kind"
>
> Thanks very much in advance to all for any info.
>
> --
> View this message in context:
http://r.789695.n4.nabble.com/Good-Decision-Trees-with-Product-Purchased-Data-tp4632438.html
> Sent from the R help mailing list archive at Nabble.com.
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.


-- 

Bert Gunter
Genentech Nonclinical Biostatistics

Internal Contact Info:
Phone: 467-7374
Website:
http://pharmadevelopment.roche.com/index/pdb/pdb-functional-groups/pdb-biostatistics/pdb-ncb-home.htm

> Define: "one" (who?) 
The individual analyzing the data.
>Define: "usually" 
More than 50% of the time.
> Define: "good" 
A decision tree including branches on at least 2 variables.

 >Define: "data of this kind" 

I believe I may have specified this in my original post:
> - ID of brand purchased 
> - Importance ratings (1-10 scale) for a number of relevant attributes 
> (price, strength, recommended by a friend, etc.)  In other words, a rating 
> of how important each attribute is in the decision as to which brand to 
> purchase. 
--
View this message in context:
http://r.789695.n4.nabble.com/Good-Decision-Trees-with-Product-Purchased-Data-tp4632438p4632444.html
Sent from the R help mailing list archive at Nabble.com.