Data Mining MCQs

A superset cannot have more support than its subset.
Because {A, B} is below threshold, any superset containing it is also below threshold.

The count records how many transactions share the represented prefix.
Links connect nodes to parents, children, and same-item occurrences.

A classifier learns a decision rule from labeled examples.
It then predicts class labels for new observations.

ROC curves can appear optimistic when true negatives dominate.
Precision-recall analysis emphasizes performance on the minority class.

Multiplying a feature by a positive constant changes threshold values but not ordering.
Distance-based domination is therefore not a concern.

Multiplying by a zero likelihood makes the entire class posterior zero.
Smoothing assigns small nonzero probabilities to unseen events.

Large neighborhoods smooth local structure and may underfit.
The majority class can overwhelm minority patterns.

Every subset of a frequent itemset must meet minimum support.
The subset may have greater support than the full itemset.

Header links allow efficient traversal of an item's occurrences across the tree.
They are used to construct conditional pattern bases.

The model learns patterns from training examples.
Performance must be assessed on separate data to estimate generalization.

Stratification reduces accidental class imbalance between partitions.
It is especially useful for small or imbalanced datasets.

A high-cardinality attribute can create many pure but tiny partitions.
Gain ratio or regularization can reduce this preference.