All MCQs

By anti-monotonicity, no superset of an infrequent set can be frequent.
Pruning avoids unnecessary support counting.

A transaction containing a larger itemset must contain all its subsets.
Therefore a superset's support is never greater.

The Apriori property follows from support anti-monotonicity.
An itemset cannot occur more often than any of its subsets.

Constraints can restrict items, rule length, aggregate values, or business conditions.
They reduce search and focus the output on relevant patterns.

Lift compares rule confidence with the consequent's baseline support.
A value above one indicates improvement over the baseline frequency.

High support reflects frequent co-occurrence in absolute terms.
Lift near one shows the co-occurrence is close to what marginal popularity predicts.

Association measures observational dependence rather than intervention effects.
Business context and causal analysis are needed before claiming causation.

Association mining can generate many overlapping rules from related itemsets.
Pruning and concise representations improve interpretability.

Different items may have very different baseline frequencies.
Multiple minimum supports or specialized methods can better handle rare but important items.

Some valuable patterns, such as rare adverse events, naturally have low support.
An excessive threshold filters them out.

Lower support allows many combinations to qualify as frequent.
This increases computation and the burden of interpreting results.

Confidence evaluates conditional rule strength.
Rules below the chosen threshold are excluded from the final set.