Dear All,
Criteria for building models
I have a question which might encompass broad
community involved in statistics. My question is about
criteria used in building multivariate statistical
models in different disciplines (to name some
Economics, Physics, Meteorology, Psychology and
Agriculture).
My question is mainly concerned with assessing
prognostic ability of multivariate models and how it
is assessed in these disciplines (for example,
measures for selecting the best set of prognostic
variables or measures for evaluating the accuracy of
predictions in each of these scientific fields).
I am familiar with measures like R-squared, C index,
Cp and some other common measures but I am interested
to know if any particular measure is used/developed
specifically in these disciplines.
Any comment is very much appreciated.
Best regards
Babak Oskooei
__________________________________
Yahoo! Music Unlimited
Access over 1 million songs. Try it free.
http://music.yahoo.com/unlimited/
|