[SOLVED] Information Gain Feature Selection with Integer Values

rvaz

Hello,

      I'm using Information Gain as a method for feature selection.
      The features represent the number of occurrence of a term in a document tf(t,d).
      My doubt is:
                         In order to calculate the IG does the method implemented in RapidMiner
                         takes in consideration all the possible values of a feature
                         or only the presence or absence of a feature in a certain example?

Thanks in Advance,

Rui Vaz

michaelhecht

Can you give an example of the features?

MariusHelf

RapidMiner considers all values, since it does not have a concept of "absence" (left aside missing values, but I think that you don't have any missing values).

rvaz

Ok thanks Marius.