Optimization

[jimjohn]

I have one variable, that is Redemption Rate ( a rate between 0 and 1).
then I have another variable that is an incentive (a value between -1 and 1)

If I calculate the correlation between these two variables, the value is .207. Now, waht I want to do is maximize the correlation between these two variables by filtering my other variables as needed. For example, maybe if I only included values of incentive > 0, then this would increase the correlation. Or, maybe if I only include Ages < 300. etc. I just want to find the right filters for my other variables (Age, original balance, Personal or commercial etc.) that will maximize this correlation between incentive and redrate.

my data file is huge, and choosing select cases, filtering the data and noticing the change in correlation again and again to come up with a max correlation will waste a lot of time. can someone plz tell me if I can somehow use SPSS to automate this procedure, and have it come up with the best filters to reach the maximum correlation. thanks so much!

[ViAnn Beadle]

Look at the TREES procedure to build a model maximizing Redemption rate. You
might find that incentive doesn't even enter the model.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
jimjohn
Sent: Friday, September 12, 2008 11:19 PM
To: [hidden email]
Subject: Optimization

I have one variable, that is Redemption Rate ( a rate between 0 and 1).
then I have another variable that is an incentive (a value between -1 and 1)

If I calculate the correlation between these two variables, the value is
.207. Now, waht I want to do is maximize the correlation between these two
variables by filtering my other variables as needed. For example, maybe if I
only included values of incentive > 0, then this would increase the
correlation. Or, maybe if I only include Ages < 300. etc. I just want to
find the right filters for my other variables (Age, original balance,
Personal or commercial etc.) that will maximize this correlation between
incentive and redrate.

my data file is huge, and choosing select cases, filtering the data and
noticing the change in correlation again and again to come up with a max
correlation will waste a lot of time. can someone plz tell me if I can
somehow use SPSS to automate this procedure, and have it come up with the
best filters to reach the maximum correlation. thanks so much!
--
View this message in context:
http://www.nabble.com/Optimization-tp19467574p19467574.html
Sent from the SPSSX Discussion mailing list archive at Nabble.com.

=====================
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

=====================
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