Association rules


  • Hello people!

    Could somebody explain the difference between the importance and the lift of a rule?

    Many thanks! :)

    Tuesday, July 23, 2013 12:08 PM

All replies

  • In data mining and association rule learning, lift is a measure of the
    performance of a model (association rule) at predicting or classifying cases as
    having an enhanced response (with respect to the population as a whole),
    measured against a random-choice targeting model.

    Importance is A measure of the usefulness of a rule. A greater value
    means a better rule.

    You can find more details to help you understand by reading my blog: Concepts in Microsoft Association Rules: Lift, Support,
    Importance, and Probability


    Thursday, August 08, 2013 8:55 PM