Communities DATA-SCIENCE-OPTIMIZE Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A data scientist is comparing two community detection algorithms on a medium-sized network. Algorithm X runs in O(n^2) time and always finds the optimal modularity partition. Algorithm Y runs in O(n log n) time and finds a good but not guaranteed optimal partition. The scientist needs to process a network that will grow to millions of nodes. Which algorithm should be chosen?

   • AEither algorithm, because both will work equally well
   • BAlgorithm Y, because it scales better and provides a good partition
   • CAlgorithm X, because optimal modularity is essential
   • DAlgorithm X, because it is more accurate
   Reveal Solution Discussion Select an answer first
2. 12expert · hard

   A data scientist is evaluating community detection results on a network with a known community structure. The scientist computes modularity, conductance, and coverage for the detected communities. Which combination of metric values would indicate the best community structure?

   • ALow modularity, low conductance, high coverage
   • BLow modularity, high conductance, low coverage
   • CHigh modularity, high conductance, low coverage
   • DHigh modularity, low conductance, high coverage
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   In the context of community detection, what does modularity measure?

   • AThe average shortest path length between nodes in the same community
   • BThe density of edges within communities relative to the expected density in a random graph with the same degree sequence
   • CThe number of connected components in the network
   • DThe number of edges inside communities compared to the number of edges between communities
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which metric measures the fraction of edges that have both endpoints inside the community, relative to the total number of edges in the network?

   • ACoverage
   • BModularity
   • CConductance
   • DDensity
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A social network analyst is studying a network of Twitter followers. The analyst runs the Louvain algorithm and finds that many users are assigned to a single large community, which is not useful for marketing segmentation. The analyst wants to obtain smaller, more focused communities. What should the analyst do?

   • AUse clique percolation with a low k value.
   • BDecrease the resolution parameter in Louvain.
   • CIncrease the resolution parameter in Louvain.
   • DSwitch to Girvan-Newman, which always produces smaller communities.
   Reveal Solution Discussion Select an answer first