Re: DM: CHAID vs. CARTŪ

[Nicolas Lachiche]

>>>>> "Mike" == Mike Muratet <muratetm@nichols.com> writes:
  > Organization: Nichols Research

  > Ronny Kohavi wrote:

  >> Since the theory says that it's impossible for one algorithm to
  >> uniformly beat any other on generalization accuracy in 
classification
  >> tasks (where uniformly is for all possible target concepts), the
  >> question is *when* (under what conditions) one algorithm is 
better
  >> than another, not *whether*.
  >> 

  > Do you have a reference to the paper or book that describes such 
a theory?
  > Thanks.

  > Mike

It seems to me the following article is a good reference on this
point:

@InProceedings{SCH94,
  author =       "Cullen Schaffer",
  title =        "A conservation law for generalization performance",
  pages =        "259-265",
  booktitle =    "Proceedings of the Eleventh International Conference
                  on Machine Learning (ICML'94)",
  year =         1994,
  publisher =    "Morgan Kaufmann",
  address =      "New Brunswick, NJ"
}

        Nicolas