hi, there:
I have a question on use of arules package:
suppose i have a classification problem with class id = 1 or 0.
i put predictors and class ids together and tranform all of them into
binary thus i got a binary matrix. When I build the rules from it, I
subset the rules to 2 by defining rhs %in% "classid = 1" or
"2".
(Actually i can build rules from two seperate datasets (by classid)
and use eclat to get most frequent itemset).
Anyway, I am thinking to use "chisq" or something to test the
differentiating power of "lhs" and look for the most powerful
"lhs".
But I just don't know how to do it. Maybe someone can explain to me
about the structure of "rules" from that package.
--
Weiwei Shi, Ph.D
Research Scientist
GeneGO, Inc.
"Did you always know?"
"No, I did not. But I believed..."
---Matrix III
