All MCQs

Association analysis identifies co-occurrence patterns such as products purchased together.
Its outputs often include frequent itemsets and implication-style rules.

Clustering is an unsupervised task that searches for natural groupings in data.
No predefined class labels are required.

Regression estimates a numeric response such as price, demand, or temperature.
It differs from classification, which predicts discrete labels.

Classification learns from labeled examples to assign observations to predefined categories.
Examples include spam versus non-spam and disease versus no disease.

Patterns must be judged for correctness, novelty, usefulness, and relevance.
A technically valid pattern may still have little practical value.

Raw data usually contain missing values, noise, and inconsistent formats.
Preprocessing prepares the data so mining algorithms can operate reliably.

KDD refers to the broader process of turning raw data into useful knowledge.
Data mining is one central step within that process.

A query answers a predefined question about stored data.
Data mining often explores the data to uncover relationships not specified in advance.

Data mining applies computational and statistical methods to reveal actionable structure in data.
Its focus is knowledge discovery rather than simple storage or reporting.

Exploratory findings can result from chance, bias, or data problems.
Transparent validation protects against overclaiming and improves scientific credibility.

Distribution displays reveal shape, while raw points and labels show the evidence and sample size.
Faceting preserves group separation and consistent comparison.

Visual alignment between two series may result from chosen scales rather than data association.
Separate panels or indexed values often support more honest comparison.