Communities DATA-SCIENCE-OPTIMIZE Practice Questions (Page 4)

Questions 16–18

1. 16application · medium

   A data scientist is using the Louvain algorithm to detect communities in a network. The scientist notices that the algorithm's result can vary between runs. What is the most likely reason for this variability?

   • AThe algorithm randomly assigns nodes to communities at the end.
   • BThe algorithm is deterministic but the network changes between runs.
   • CThe algorithm uses a random null model that changes each run.
   • DThe algorithm is stochastic and depends on the order of node processing.
   Reveal Solution Discussion Select an answer first
2. 17expert · hard

   A data scientist is working with a network that has a known ground-truth community structure. The scientist wants to evaluate how well a community detection algorithm recovers the true communities. Which metric is most appropriate?

   • AConductance
   • BModularity
   • CCoverage
   • DNormalized mutual information (NMI)
   Reveal Solution Discussion Select an answer first
3. 18foundation · easy

   How do modularity optimization algorithms, such as the Louvain method, find communities?

   • ABy assigning each node a random community label and then updating labels based on neighbor majority
   • BBy removing edges with the highest betweenness centrality until the network splits into communities
   • CBy finding all maximal cliques and merging overlapping ones to form communities
   • DBy iteratively moving nodes between communities to increase modularity, then aggregating the network and repeating
   Reveal Solution Discussion Select an answer first