Re: CCL:Correlation Coefficient

On Thu, 2005-04-14 at 12:48 +0800, lhhu () yangtze.hku.hk wrote:
 > Dear All,
 >
 > I have a question about Correlation Coefficient of cross-validation. Can
 > anyone tell me how much the value of Correlation Coefficient indicate
 > the result is reliable? Certainly I know the larger the value, the more
 > reliable the result. I want to know the lowest limit of that value. I read
 > some papers that used the Correlation Coefficient value, some values are
 > lower than 0.8, is it OK for present them in journals? Or it depends on
 > different situation or topic?
 There are differing views on the validity of q^2, in that high values of
 q^2 do not necessarily indicate a good model in terms of predictive
 ability. Some papers that discuss this topic are
  (+ at +) ARTICLE{q2-1,
     author = {Golbraikh, A. and Tropsha, A.},
     title = {Beware of $q^2$},
     journal = {J.~Mol.~Graph.~Model.},
     year = {2002},
     volume = {20},
     pages = {269--276},
 }
  (+ at +) ARTICLE{q2-5,
     author = {Norinder, U.},
     title = {Single and domain made variable selection in {3D QSAR}
         applications},
     journal = {J.~Chemomet.},
     year = {1996},
     volume = {10},
     pages = {95--105},
 }
  (+ at +) ARTICLE{q2-2,
     author = {Golbraikh, A. and Shen, M. and Xiao, Z.Y. and Xiao, Y.D.
         and Lee, K.H. and Tropsha, A.},
     title = {Rational selection of training and test sets for the
         development of validated {QSAR} models},
     journal = {J.~Comput.~Aid.~Mol.~Des.},
     year = {2003},
     volume = {17},
     pages = {241--253},
 }
 HTH
 -------------------------------------------------------------------
 Rajarshi Guha <rxg218 () psu.edu> <http://jijo.cjb.net>;
 GPG Fingerprint: 0CCA 8EE2 2EEB 25E2 AB04 06F7 1BB9 E634 9B87 56EE
 -------------------------------------------------------------------
 I'm related to people I don't relate to."
 -Calvin