DM: AW: Classification problem

If you are working with a standard DM Software there are two way:
1. duplicate the bad guy 5000times
2. select only the most important examples of the good guys. (or a mix of 1
and 2 )

Because all this isn't quite practical, the best method is to tackle on the
error function of the learning algorithm. Weight the error produced buy
misclassification of bad guys 5000 times harden than the error of the good
guys.

Therefore you need to have inside in the algo. That's the advantage of DM
Toolboxes for Software like MATLAB.

Note that the extent of duplicating or weighting is an expression of how
many costs would an misclassification in practice produce.

Frank Buckler ------------------------------------------------------------
University of Hanover
Dep. Marketing II
buckler@m2.uni-hannover.de

-----Ursprüngliche Nachricht-----
Von: owner-datamine-l@nautilus-sys.com
[mailto:owner-datamine-l@nautilus-sys.com]Im Auftrag von Yannis Kopanas
Gesendet: Donnerstag, 25. Mai 2000 06:17
An: datamine-l
Betreff: DM: Classification problem


My problem has to do with the data set. I have two classes (the good guys
and the bad guys) unfortunatelly the bad guys are only 20 when the good guys
are 99980. Anybody who knows how to deal with it?
Thanks in advance.
      Yannis

• Prev by Date: DM: RE: Data Forms for Mining (Limit on variables)
• Next by Date: DM: Final call for papers: Application of AI in industry
• Prev by thread: DM: Final call for papers: Application of AI in industry
• Next by thread: Re: DM: AW: Classification problem
• Index(es):
  • Date
  • Thread