In an association rule, which metric measures the probability of the consequent given the antecedent?

• A. Confidence
• B. Support
• C. Lift
• D. Conviction

Answer: (A)

Think of an association rule as A implies B. The question asks for the metric that captures how likely B is when A has occurred. That is exactly what confidence measures: the probability of the consequent given the antecedent, denoted P(B|A). In practical terms, confidence equals the frequency of transactions containing both A and B divided by the frequency of transactions containing A.

Support, on the other hand, is the joint frequency P(A and B) and doesn’t reflect conditioning on A. Lift compares P(B|A) to P(B) to see if A and B occur together more often than by chance. Conviction relates to how often A occurs without B and provides a different angle on implication strength.