Type your query or search by category
< All Topics

Which algorithm should I use, K-means or G-means?

AI Studio offers two different algorithms for clustering: K-means and G-means. Both algorithms group the most similar instances in your dataset. The difference between them is how they accomplish the pipeline.

  • With K-means you need to select the number of clusters to create. You can decide how each field in your dataset influences which group each instance belongs to. K-means is the default algorithm when you select CONFIGURE CLUSTER from the configure option menu. AI Studio creates eight clusters and applies automatic scaling to all the numeric fields. The number of clusters, weighting, as well as field scaling and can easily be modified when you configure your cluster from the AI Studio Canvas. K-means is useful when you already know the number of clusters you will get from your dataset when grouping all your instances.
Table of Contents