To build the decision tree, we use the Gini impurity or information gain to determine the threshold for each node in this tree. my question is that if there is any possible parameter like information gain or Gini impurity could build the single split node for the tree? Thank you