Which parameter shortens runtime when clustering large datasets with k-Means in RapidMiner?

• A. Number of clusters
• B. Tolerance
• C. Max runs
• D. Seed

Answer: (C)

In k-Means, runtime grows with how much work the algorithm does across iterations and restarts. Each run repeatedly assigns points to the nearest centroid and then updates centroids. If you perform multiple restarts with different initial centroids, you multiply that workload. RapidMiner’s max runs parameter controls how many such restarts the operator will perform. By limiting this number, you reduce the total number of clustering attempts, cutting down distance calculations and centroid updates, which speeds up processing on large datasets. The other factors influence runtime in less direct ways: more clusters increases per-iteration work and can slow things down; tolerance affects when a single run converges; seed only changes starting points and doesn’t usually have a large impact on overall runtime.