Return to Design view in RapidMiner. Delete the operators in the scoring stream. Connect the exa port from the Decision Tree operator to the unl port on the Apply Model operator. Add a Performance $($Classification$)$ operator to your process. Connect the lab ports between Apply Model and Performance. Output the model$$s performance $($per$)$ to a result set $($res$).$ Accept the Performance operator$$s default parameters.

Using the criterion parameter for the Decision Tree operator, run the process four times, once for each of the first four criterion algorithm options: $$gain$\_$ratio$,$information$\_$gain$,$gini$\_$index$,$ and $$accuracy$.$ Note each algorithm$$s Performance Vector accuracy percentage. Of the four algorithms, which had the highest overall predictive accuracy for this exercise$$s training data?

accuracy

gini$\_$index

gain$\_$ratio

information$\_$gain