Clustering in Data Workbench

[james_o_hagan]

Hi all, 

     The company I work at is getting Data Workbench and I was reading up on the clustering options in the predictive analytics section and had a few questions on it. 

1. The clustering option uses k-means to partition your data set correct? So why isn't this considered a descriptive analytics option ? Is there some feature that makes predictions for each of these clusters that's not in the documentation

2. A critical part of k-means clustering is normalization what type of normalization goes on when someone uses this product?

Full disclosure I haven't actually used the product yet so I'm sure there are a lot of features that I haven't looked at yet that might warrant  grouping in the clustering tool with the other predictive analytics tools.

Best, 

       James O'Hagan

[TanmayMathur · Accepted solution]

Hi James,

1. The clustering feature is part of the predictive workflow made available with Analytics Premium. To your point, it is a descriptive component of that workflow.

2. Users can choose between two normalization methods. Either the Min-Max or the Z-Score technique. The Min-Max method is enabled by default.

I hope that helps.