RapidMiner 9.7 is Now Available

Lots of amazing new improvements including true version control! Learn more about what's new here.

CLICK HERE TO DOWNLOAD

Association Rules: Infinity for Lift and Confidence

btibertbtibert Member, University Professor Posts: 80  Guru
edited October 2019 in Help
Admittedly I am stumped as to why every single rule in my dataset has a value of infinity for lift.  What could the intuition of this possibly be?



If it helps, I have attached my dataset.

And the settings




Comments

  • lionelderkrikorlionelderkrikor Moderator, RapidMiner Certified Analyst, Member Posts: 1,074   Unicorn
    Hi @btibert,

    Unfortunately, I'm not an expert in "association rules" theory, but I found an answer element in the community : 

    https://community.rapidminer.com/discussion/50468/fp-growth-produced-infinite-value-for-association-rules

    Hope this helps,

    Regards,

    Lionel
  • IgorEIgorE Member Posts: 4 Learner I
    I have a similar question about infinite confidence. 
    conf(X implies Y) = supp(X ∪Y) / supp(X) .
     supp(X ∪Y) = occurs(X ∪Y) / total_records
     supp(X) = occurs(X) / total_records

    so (if I'm not mistaken)

     conf(X implies Y) = occurs(X ∪Y) / occurs(X)

    How can it amount to infinity?

    I have two very similar datasets. RM generates very similar rules but in one case for the same rule, the confidence is 0.99 in another it's infinity.

    Please advise,
    Igor
Sign In or Register to comment.