How to separate 10 attributes with Rapidminer?

Question

I have an excel file with a lot of data.I want to find 10 columns where cause best separation. I want to do this with rapid miner. (sorry my English is not good)

Answer 1

You can use one of the Weight By operators such as Weight By Correlation, to create a set of attribute weights where the weight of an attribute represents how predictive it is of the class label. You can then use the Select By Weight operator to select the desired number of attributes. Set the parameters of this operator to weight relation to top k and k to 10. The inputs to this operator are the attribute weights and the example set which are output by the Weight By operator.