Clustering GMLE Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   Which of the following is an example of an unsupervised learning task?

   • AEstimating the probability of a patient readmission
   • BGrouping customers into segments based on purchasing behavior
   • CPredicting house prices from square footage and location
   • DClassifying emails as spam or not spam
   Reveal Solution Discussion Select an answer first
2. 2expert · hard

   A geospatial analytics company clusters millions of GPS points to identify traffic congestion zones. The data has varying densities: urban areas have high density, while rural highways have lower density. They need to identify clusters of arbitrary shape and also distinguish noise from sparse traffic. They tried DBSCAN with a fixed epsilon, but it either merges urban clusters or misses rural ones. Which approach is most appropriate?

   • AUse K-Means with a high k value
   • BUse DBSCAN with a very small epsilon
   • CUse OPTICS with a maximum epsilon and minPts
   • DUse hierarchical clustering with single linkage
   Reveal Solution Discussion Select an answer first
4. 3expert · hard

   A company wants to cluster customer feedback text to identify common issues. They have a large corpus of reviews. They plan to use K-Means on TF-IDF vectors. The initial clustering produces clusters that are not interpretable, and the silhouette score is low. They suspect the high dimensionality and sparsity are causing the problem. Which combination of preprocessing and algorithm changes is most likely to improve the clustering?

   • AUse K-Means with Euclidean distance on the raw TF-IDF matrix
   • BApply Truncated SVD to reduce dimensions, then use K-Means with cosine distance
   • CUse DBSCAN with a small epsilon on the raw TF-IDF matrix
   • DUse hierarchical clustering with single linkage on the raw TF-IDF matrix
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which method is commonly used to help select the optimal number of clusters for K-Means?

   • AThe gradient descent method
   • BThe bootstrap method
   • CThe one-hot encoding method
   • DThe elbow method
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   In DBSCAN, what does the parameter 'epsilon' (eps) control?

   • AThe maximum distance between two points to be considered neighbors
   • BThe minimum number of points required to form a cluster
   • CThe maximum number of clusters to find
   • DThe number of iterations to run
   Reveal Solution Discussion Select an answer first