Dear Sir
Hi. In order to validate a regression model, it is better to obtain some
statistics (for example: AIC, Cp, SBC, and so on) from both model-building
set and validation set (cross-validation process). The statistical package
that I use is SPSS 17.0 Professor Package, and OS Windows xp Pro. I would
like to make a cross-validation to test a general linear regression model
with two variables. The following table has been written in a textbook.
Would you mind email me, please, how I can create a table like this:
Kind regards
Mehrshad
Model 1 Model 2
Training Data Set Validation Data Set
P a real number a real number
b0 a real number a real number
b0 sd a real number a real number
b1 a real number a real number
b1 sd a real number a real number
b2 a real number a real number
b2 sd a real number a real number
SSEp a real number a real number
PRESSp a real number a real number
Cp a real number a real number
MSEp a real number a real number
MSPR a real number a real number
Ra,b2 a real number a real number
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
