In the preceding recipe, we saw how to use ctree
for control inference trees. We can tweak the algorithm by specifying the control parameters.
Perform the following steps in R:
> ctree.model = ctree(churn ~ . , data = trainset, controls=ctree_control(testtype = "MonteCarlo", mincriterion = 0.90, minbucket = 15))> ctree.model
This recipe is a continuation of the previous recipe, providing some control parameters using a controls argument to the ctree
function. We said we are going to use the MonteCarlo simulation with minimum weight on node at 15 and mincriterion at 0.90. We can also use Bonferroni, Univariate, and Teststatistic in place of MonteCarlo. There are many other parameters that can be changed.