Question

Do you have to remove variables with high standard deviation for a decision tree on sas miner? If you do not what are the consequences?

Answer 1

Do you have to remove variables with high standard deviation for a decision tree on sas miner? If you do not what are the consequences?

Answer:

Standard Deviation — standard deviation values for numeric variables only. For character variables, “.” (missing notation) is displayed.

In the SAS Miner SEMMA data mining methodology, the Drop node belongs to the Modify category. You use the Drop node to remove variables from data sets or hide variables from the metadata. You can remove all variables with the role type that you specify in the Drop node Properties Panel, or you can manually specify individual variables to drop using the “Drop Node Variables Table”. For example, you could remove all hidden, rejected, and residual variables from your exported data set, or just a few variables that you identify yourself.