statistical testing

Debasish Roy d.roy at auckland.ac.nz
Thu Feb 24 19:12:27 PST 2005


(Apologies for cross posting)

Dear Colleagues:

There are 2 bunches of samples and they have density functions (or,
distribution functions) say, F and G.  But the samples are DEPENDENT and
they originated from MCMC simulation. Now I want to test H_0 : F(.)  =
G(.)

One imporatnt thing to note that we know F(.) is normal. But we have not
good idea of the true density function of G(.). And also it's not
possible to find the joint parametric density of that 2 datasets.

I can't see any hope to find the suitable test as the samples are
dependent. So standard non-para tests fail. Could you advise me the best
test here ? How about Kalman Filtering ? Though I don't know much about
it.

I look forward to hearing your kind suggestion at your earliest
opportunity.


Regards
Debasish.



-------------
The compgeom mailing lists: see
http://netlib.bell-labs.com/netlib/compgeom/readme.html
or send mail to compgeom-request at research.bell-labs.com with the line:
send readme
Now archived at http://www.uiuc.edu/~sariel/CG/compgeom/maillist.html.



More information about the Compgeom-announce mailing list