An IS auditor reviewing documentation for an AI model notes that the modeler utilized a K-means clustering algorithm, which clusters data into categories for correlations and analysis. Which of the fo

Question

Options

• AK-means clustering is not a common data clustering method due to its complexity and difficulty
• BK-means clustering requires the modeler to supervise the learning analysis, which can introduce
• CK-means clustering algorithms are significantly sensitive to outliers and dependent on the similarity
• DK-means clustering determines the number of clusters for the modeler without supervision.

• A
  3% (1)
• B
  13% (5)
• C
  77% (30)
• D
  8% (3)

Explanation

K-means clustering is an unsupervised algorithm that groups data based on similarity (typically Euclidean distance). Its critical weakness is sensitivity to outliers-a single extreme data point can significantly skew cluster centroids-and its results are heavily dependent on the chosen distance metric and the initial placement of centroids. This means model outputs can be misleading or unstable depending on data quality and configuration choices. Option A is false: K-means is widely used and straightforward to implement. Option B is false: K-means is unsupervised, meaning no labeled supervision is required. Option D is false: K-means requires the modeler to pre-specify the number of clusters (k); it does not determine this automatically.

No community discussion yet for this question.