1. Misclassification error the node A:

2. Misclassification error the node B:

3. Information gain in Misclassification error:

Suppose that you are developing a decision tree where the root node (parent node) has 50 data points with 20 corresponding to "Yes" and 30 corresponding to "No" class of the target variable. You split the tree based on an input variable (which has two categories) and create two branches, A and B (see the figure below). In this split. Branch B has 20 data points out of which 18 belong to "Yes" and 12 belong to "No" class of the target variable