When evaluating all possible splits

$1$ point

of a decision tree what can be used to find the best split regardless of what happened in prior or future steps?

Greedy Search

Classification

Regularization

Logistic regression