Typo?
class weight is not so much a metric, but more of a parameter.Consider a binary classification problem where the dataset has 1000 of 1's and 50 of 0's. Which of the metric from the DecisionTreeClassifier is used in order to adjust for the data inconsistency?