| Date: | Thu, 12 Jan 2012 08:29:48 -0600 |
|---|
| Reply-To: | "Farmaha, Bhupinder" <bhupi80singh@YAHOO.CO.IN> |
|---|
| Sender: | "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU> |
|---|
| From: | "Farmaha, Bhupinder" <bhupi80singh@YAHOO.CO.IN> |
|---|
| Subject: | Re: Clarification on Concept of k-fold Cross Validation |
|---|
|
| In-Reply-To: | <201201121420.q0C5cYOU009815@waikiki.cc.uga.edu> |
|---|
| Content-Type: | text/plain; charset="us-ascii" |
|---|
Rick,
Would it be possible for compare linear versus non-linear MIXED models?
Thanks
Bhupinder
-----Original Message-----
From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of Rick
Wicklin
Sent: Thursday, January 12, 2012 8:20 AM
To: SAS-L@LISTSERV.UGA.EDU
Subject: Re: Clarification on Concept of k-fold Cross Validation
You don't use any of the k models. You use the model evaluated on the full
data.
Cross-validation is useful for comparing several models. The cross-
validation method estimates the prediction error for each model. You can use
cross-validation (and AIC and BIC....) to help you decide which of several
candidate models you want to choose to be THE model for the data.
Details at:
http://support.sas.com/documentation/cdl/en/statug/63033/HTML/default/viewe
r.htm#statug_glmselect_sect025.htm
The second paragraph is especially helpful.
Rick Wicklin
| 
