I've been playing a bit with clustering and came to the situation where I would like to be able to use the cluster centers of a cluster model as example set (I would basically like to feed the clusters centers to a previously built decision tree). Unfortunately I couldn't find any operator that allows me to do so. Is there an automatic way to convert cluster centers to an example set?
1) cluster the data 2) divide the data into groups according to the cluster name, e.g. with the ValueIterator + ExampleFilter 3) calculate the mean values for all features and store them by using the process log 4) build an example set from the process log table
Just a tip: be creative! I often think "oh, I am missing an operator here" but then after a small period of trying I always managed to get things done by using a more or less complex combination of the standard operators.