Regression: Best Subsets

[trawl13]

Hi,

I need help finding the the most significant, non-correlated, variables in SAS.

The Current Approach:

In the "Analysis" section, I am currently going line by line to find which variables are correlated. I have 187 variables, so that is 2^187 (=1.961e56) combinations I have to go through. I am using the Pearson Correlation Coefficient to find multicolinearity b/t variables. I am eliminating ones that are >.80

I know that SAS can hadle the best subsets, but I would like the best subsets for 187 variables taking into accounf multicolinearity

Ideally I would like to:

1.) Eliminate all of the correlation b/t the 187 variables

2.) With the non correlated variables, I would like to find the most significant, best subset of variables to run in a regression

This is taking too long, and I know there is a more mathematical approach to solve this. That is why I am reaching out to people much smarter than myself

Thanks for any help!

[Reeza]

Have you looked into Principal Compenents Analysis (PCA)?

[PGStats]

This is a vast subject, full of traps. Here is one approach. You should try doing a PCA (proc princomp) to determine roughly how many true independent factors are among your variables. Then you should look for best subsets of that size using best subset regression. Last, you should check, starting with the bestest subset, that the chosen variables are indeed uncorrelated (it is unlikely that correlated variables will turn out in the same subset). After all that, make sure the model makes sense, that it is meaningful!

My two cents.

PG