Re: Regression of correlated variables

Tim vor der Brück wrote:
David Jones wrote:
Tim vor der Brück wrote:
Hi,

I would like to do a regression of highly correlated variables.
Could somebody give me a hint about the best method. Principle
Component Analysis?

Thank your your any help,

Tim

You may need to be a little more specific about what you mean. In
particular, if you have a set of correlated variables that you are
trying to predict from another set of correlated variables, then
what
you want might be "Cannonical Correlations" rather than PCA. Ojn
the
other hand you may have a pure multiple regression problem, in
which
case you need to say why you think a direct approach won't work for
you ... is it a question of selecting which "explanatory variables"
to include.

David Jones


Hi, thank you for your answer.

I have a problem y=a1*x1+...+an*xn
x1,...,xn are correlated with each other.

Tim vor der Brück

This appears to be just an ordinary multiple regression problem,
unless there are other features you haven't states.

David Jones


.

Relevant Pages