Data Mining MCQs

Compatible (k-1)-itemsets are united to create k-itemset candidates.
Prefix ordering prevents many duplicates.

Missingness means the true value is not present in the dataset.
It should not automatically be interpreted as zero.

This method reduces local noise by representing nearby values with a common summary.
It also sacrifices within-bin detail.

Frequency is determined by the user-defined minimum support.
Only qualifying itemsets are used to generate standard association rules.

Every maximal frequent itemset is closed, but not every closed itemset is maximal.
Closed sets retain more support information.

A candidate with an infrequent subset cannot be frequent.
This follows directly from the Apriori property.

Under MCAR, missing records form an unbiased subset with respect to the variables.
This is a strong assumption and often unrealistic.

Outliers may reflect error, rare events, or legitimate extremes.
They should be investigated rather than automatically removed.

The antecedent X implies an increased occurrence of consequent Y.
The two sides must not share items in a standard rule.

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

Each level typically generates new candidates and counts their occurrences.
This scan-intensive behavior can be expensive for large datasets.

The mean can be unrepresentative for skewed data.
Mean imputation also reduces variance and can distort relationships.